[
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "moment-1",
  "level": "related",
  "reason": "related dataset Weather (Jena, 2020; Autoformer) was used (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "timer-s1",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "units-few-shot-anomaly-detection",
  "level": "related",
  "reason": "via base model UniTS: related dataset Weather (Jena, 2020; Autoformer) was used",
  "sources": "https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "jena-weather",
  "model": "units",
  "level": "related",
  "reason": "related dataset Weather (Jena, 2020; Autoformer) was used",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "moment-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "timer-s1",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-application",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "moment-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "timer-s1",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-service",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "moment-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "timer-s1",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bizitobs-l2c",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "moment-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "timer-s1",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-fast-storage",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "moment-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "timer-s1",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "bitbrains-rnd",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "moirai-1.0-r",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "moirai-moe-1.0-r",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "moment-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "timer-s1",
  "level": "train_seen",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive)); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "toto-2024",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "recruit-restaurant",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "lag-llama",
  "level": "related",
  "reason": "related dataset ETTh1 was used",
  "sources": "https://arxiv.org/pdf/2310.08278v3 https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2205.13504v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "moment-1",
  "level": "related",
  "reason": "related dataset ETTh1 was used (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2205.13504v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "timer-s1",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "units-few-shot-anomaly-detection",
  "level": "related",
  "reason": "via base model UniTS: related dataset ETTh1 was used",
  "sources": "https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2205.13504v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett1",
  "model": "units",
  "level": "related",
  "reason": "related dataset ETTh1 was used",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2205.13504v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "lag-llama",
  "level": "related",
  "reason": "related dataset ETTh2 was used",
  "sources": "https://arxiv.org/pdf/2310.08278v3 https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2205.13504v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "moment-1",
  "level": "related",
  "reason": "related dataset ETTh2 was used (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2205.13504v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "timer-s1",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "ett2",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2401.03955v8"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2401.03955v8"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "moirai-1.0-r",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "moirai-moe-1.0-r",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "moment-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "timer-s1",
  "level": "train_seen",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive)); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "toto-2024",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "ttm-research-r2",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2401.03955v8"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "loop-seattle",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2401.03955v8"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2401.03955v8"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "moirai-1.0-r",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "moirai-moe-1.0-r",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "moment-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "timer-s1",
  "level": "train_seen",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive)); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "toto-2024",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "ttm-research-r2",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2401.03955v8"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "sz-taxi",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "moirai-1.0-r",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "moirai-moe-1.0-r",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive)); this contradicts a statement that the data was excluded",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "moment-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4 https://huggingface.co/api/datasets/Maple728/Time-300B"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "timer-s1",
  "level": "train_seen",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive)); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "toto-2024",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m-dense",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "chronos-2",
  "level": "related",
  "reason": "related dataset NREL Solar Power Data for Integration Studies (2006) was used; the authors state this dataset's test portion was excluded",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://zenodo.org/records/4656144 https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "chronos-bolt",
  "level": "related",
  "reason": "related dataset NREL Solar Power Data for Integration Studies (2006) was used (via Chronos training corpus (real-world datasets))",
  "sources": "https://github.com/amazon-science/chronos-forecasting/issues/245 https://arxiv.org/pdf/2403.07815v3 https://zenodo.org/records/4656144"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "chronos-t5",
  "level": "related",
  "reason": "related dataset NREL Solar Power Data for Integration Studies (2006) was used (via Chronos training corpus (real-world datasets))",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://zenodo.org/records/4656144"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "granite-ttm-r1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "granite-ttm-r2-1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "granite-ttm-r2",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "lag-llama",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "moment-1",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "timer-s1",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "tirex-2-fevbench",
  "level": "related",
  "reason": "related dataset NREL Solar Power Data for Integration Studies (2006) was used",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://zenodo.org/records/4656144 https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "tirex-2-gifteval-zs",
  "level": "related",
  "reason": "related dataset NREL Solar Power Data for Integration Studies (2006) was used; the authors state this dataset was excluded",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://zenodo.org/records/4656144 https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "tirex",
  "level": "related",
  "reason": "related dataset NREL Solar Power Data for Integration Studies (2006) was used",
  "sources": "https://arxiv.org/pdf/2505.23719v2 https://zenodo.org/records/4656144 https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "tspulse-r1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2505.13033v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "ttm-research-r2",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2401.03955v8"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "lstnet-solar",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "granite-ttm-r2-1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "moirai-1.0-r",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "moirai-moe-1.0-r",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "moment-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "timer-s1",
  "level": "train_seen",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive)); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "toto-2024",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hierarchical-sales",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "moirai-1.0-r",
  "level": "test_seen",
  "reason": "the whole dataset was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "moirai-moe-1.0-r",
  "level": "test_seen",
  "reason": "the whole dataset was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "moment-1",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "timer-s1",
  "level": "train_seen",
  "reason": "the whole dataset was used (via LOTSA (Large-scale Open Time Series Archive)); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "timer-xl",
  "level": "test_seen",
  "reason": "the whole dataset was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "timer",
  "level": "test_seen",
  "reason": "the whole dataset was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "timesfm-1-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "timesfm-2-0",
  "level": "test_seen",
  "reason": "the whole dataset was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "toto-2024",
  "level": "test_seen",
  "reason": "the whole dataset was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-yearly",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "moirai-1.0-r",
  "level": "test_seen",
  "reason": "the whole dataset was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "moirai-moe-1.0-r",
  "level": "test_seen",
  "reason": "the whole dataset was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "moment-1",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "timer-s1",
  "level": "train_seen",
  "reason": "the whole dataset was used (via LOTSA (Large-scale Open Time Series Archive)); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "timer-xl",
  "level": "test_seen",
  "reason": "the whole dataset was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "timer",
  "level": "test_seen",
  "reason": "the whole dataset was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "timesfm-1-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "timesfm-2-0",
  "level": "test_seen",
  "reason": "the whole dataset was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "toto-2024",
  "level": "test_seen",
  "reason": "the whole dataset was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-quarterly",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "chronos-2",
  "level": "train_seen",
  "reason": "the dataset was used; the source does not say which part; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "chronos-bolt",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))",
  "sources": "https://github.com/amazon-science/chronos-forecasting/issues/245 https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "chronos-t5",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "moirai-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "moirai-moe-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "moment-1",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "timer-s1",
  "level": "train_seen",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive)); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "timesfm-1-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "tirex",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "toto-2024",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-monthly",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "chronos-2",
  "level": "train_seen",
  "reason": "the dataset was used; the source does not say which part; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "chronos-bolt",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))",
  "sources": "https://github.com/amazon-science/chronos-forecasting/issues/245 https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "chronos-t5",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "moirai-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "moirai-moe-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "moment-1",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "timer-s1",
  "level": "train_seen",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive)); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "timesfm-1-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "tirex",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "toto-2024",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-weekly",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "chronos-2",
  "level": "train_seen",
  "reason": "the dataset was used; the source does not say which part; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "chronos-bolt",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))",
  "sources": "https://github.com/amazon-science/chronos-forecasting/issues/245 https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "chronos-t5",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "moirai-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "moirai-moe-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "moment-1",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "timer-s1",
  "level": "train_seen",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive)); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "timesfm-1-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "tirex",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "toto-2024",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-daily",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "chronos-2",
  "level": "train_seen",
  "reason": "the dataset was used; the source does not say which part; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "chronos-bolt",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))",
  "sources": "https://github.com/amazon-science/chronos-forecasting/issues/245 https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "chronos-t5",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "moirai-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "moirai-moe-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "moment-1",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "timer-s1",
  "level": "train_seen",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive)); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "timesfm-1-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "tirex",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "toto-2024",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "m4-hourly",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "moirai-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "moirai-moe-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "moment-1",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "timer-s1",
  "level": "train_seen",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive)); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "toto-2024",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "hospital",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "granite-ttm-r2-1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "moirai-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "moirai-moe-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "moment-1",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "timer-s1",
  "level": "train_seen",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive)); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "toto-2024",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "covid-deaths",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "dada",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2405.15273v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "granite-ttm-r1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "granite-ttm-r2-1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "granite-ttm-r2",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "moirai-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "moirai-moe-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "moment-1",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "sundial",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD)); this contradicts a statement that the data was excluded",
  "sources": "https://huggingface.co/thuml/sundial-base-128m/raw/main/README.md https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "timer-s1",
  "level": "train_seen",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive)); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "timeradar",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2602.19068v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "toto-2024",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "tspulse-r1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2505.13033v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "ttm-research-r2",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2401.03955v8"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "us-births",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "dada",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2405.15273v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "granite-ttm-r1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "granite-ttm-r2-1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "granite-ttm-r2",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "moirai-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "moirai-moe-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "moment-1",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "sundial",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD)); this contradicts a statement that the data was excluded",
  "sources": "https://huggingface.co/thuml/sundial-base-128m/raw/main/README.md https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "timer-s1",
  "level": "train_seen",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive)); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "timeradar",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2602.19068v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "toto-2024",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "tspulse-r1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2505.13033v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "ttm-research-r2",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2401.03955v8"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "saugeen-river-flow",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "chronos-2",
  "level": "train_seen",
  "reason": "the dataset was used; the source does not say which part; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "chronos-bolt",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))",
  "sources": "https://github.com/amazon-science/chronos-forecasting/issues/245 https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "chronos-t5",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "granite-ttm-r2-1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "moirai-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "moirai-moe-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "moment-1",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "sundial",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD)); this contradicts a statement that the data was excluded",
  "sources": "https://huggingface.co/thuml/sundial-base-128m/raw/main/README.md https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "timer-s1",
  "level": "train_seen",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive)); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "tirex",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "toto-2024",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "temperature-rain",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "chronos-2",
  "level": "train_seen",
  "reason": "the dataset was used; the source does not say which part; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "chronos-bolt",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))",
  "sources": "https://github.com/amazon-science/chronos-forecasting/issues/245 https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "chronos-t5",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "dada",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2405.15273v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "granite-ttm-r1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "granite-ttm-r2-1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "granite-ttm-r2",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "lag-llama",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "moirai-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "moirai-moe-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "moment-1",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "sundial",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD)); this contradicts a statement that the data was excluded",
  "sources": "https://huggingface.co/thuml/sundial-base-128m/raw/main/README.md https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "timer-s1",
  "level": "train_seen",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive)); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "timeradar",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2602.19068v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "tirex",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "toto-2024",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "tspulse-r1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2505.13033v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "ttm-research-r2",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2401.03955v8"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "kdd-cup-2018",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "moirai-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "moirai-moe-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "moment-1",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "timer-s1",
  "level": "train_seen",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive)); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "toto-2024",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "car-parts",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "chronos-2",
  "level": "train_seen",
  "reason": "the dataset was used; the source does not say which part; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "chronos-bolt",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))",
  "sources": "https://github.com/amazon-science/chronos-forecasting/issues/245 https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "chronos-t5",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "granite-ttm-r3",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "lag-llama",
  "level": "related",
  "reason": "related dataset Electricity (ECL, 321 clients) was used",
  "sources": "https://arxiv.org/pdf/2310.08278v3 https://raw.githubusercontent.com/laiguokun/multivariate-time-series-data/master/README.md https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "moirai-2.0-r",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "moment-1",
  "level": "related",
  "reason": "related dataset Electricity (ECL, 321 clients) was used (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE https://raw.githubusercontent.com/laiguokun/multivariate-time-series-data/master/README.md https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "timer-s1",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/Timer-S1/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "timesfm-1-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "tirex-1.1-gifteval",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "tirex-2-gifteval-pretrain",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://huggingface.co/NX-AI/TiRex-2-gifteval-zs/raw/main/README.md https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "tirex",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "toto-1.0",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "toto-2.0-2.5b-ft",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2 https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "ttm-r3",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/SalesforceAIResearch/gift-eval/main/results/TTM-R3-PT/config.json"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "units-few-shot-anomaly-detection",
  "level": "related",
  "reason": "via base model UniTS: related dataset Electricity (ECL, 321 clients) was used",
  "sources": "https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py https://raw.githubusercontent.com/laiguokun/multivariate-time-series-data/master/README.md https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "gift-eval",
  "dataset": "uci-electricity-load-diagrams",
  "model": "units",
  "level": "related",
  "reason": "related dataset Electricity (ECL, 321 clients) was used",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py https://raw.githubusercontent.com/laiguokun/multivariate-time-series-data/master/README.md https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "aa-clip-mvtec",
  "level": "seen_split_unclear",
  "reason": "in fine-tuning data: the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2503.06661v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "aa-clip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "adaclip-all",
  "level": "test_seen",
  "reason": "in fine-tuning data: its test split was used for training",
  "sources": "https://raw.githubusercontent.com/caoyunkang/AdaCLIP/main/README.md https://raw.githubusercontent.com/caoyunkang/AdaCLIP/main/dataset/base_dataset.py"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "adaclip-mvtec-colondb",
  "level": "test_seen",
  "reason": "in fine-tuning data: its test split was used for training",
  "sources": "https://raw.githubusercontent.com/caoyunkang/AdaCLIP/main/README.md https://raw.githubusercontent.com/caoyunkang/AdaCLIP/main/dataset/base_dataset.py"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "adaclip-visa-clinicdb",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "adaptclip-mvtec",
  "level": "test_seen",
  "reason": "in fine-tuning data: its test split was used for training",
  "sources": "https://arxiv.org/pdf/2505.09926v2"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "adaptclip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "anomaly-ov",
  "level": "seen_split_unclear",
  "reason": "in fine-tuning data: the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2502.07601v2"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "anomalyclip-mvtec",
  "level": "test_seen",
  "reason": "in fine-tuning data: its test split was used for training",
  "sources": "https://arxiv.org/pdf/2310.18961v12"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "anomalyclip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "anomalydino",
  "level": "unknown",
  "reason": "unknown for its base model DINOv2: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "anomalygpt-mvtec",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2308.15366v4"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "anomalygpt-supervised",
  "level": "seen_split_unclear",
  "reason": "in fine-tuning data: the dataset was used; the source does not say which part",
  "sources": "https://raw.githubusercontent.com/CASIA-IVA-Lab/AnomalyGPT/main/README.md https://raw.githubusercontent.com/CASIA-IVA-Lab/AnomalyGPT/main/code/datasets/all_supervised_with_cn.py"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "anomalygpt-visa",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "april-gan-mvtec",
  "level": "test_seen",
  "reason": "in fine-tuning data: its test split was used for training",
  "sources": "https://arxiv.org/pdf/2305.17382v3"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "april-gan-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "clip-openai",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "dinov2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "dinov3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "filo-mvtec",
  "level": "test_seen",
  "reason": "in fine-tuning data: its test split was used for training",
  "sources": "https://arxiv.org/pdf/2404.13671v3"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "filo-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "imagenet-wrn50",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/pytorch/vision/main/torchvision/models/resnet.py"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "inctrl-mvtec",
  "level": "test_seen",
  "reason": "in fine-tuning data: the whole dataset was used",
  "sources": "https://arxiv.org/pdf/2403.06495v3"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "inctrl-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenCLIP (LAION-400M checkpoints): training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "metauas",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "musc",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "openclip-laion2b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "openclip-laion400m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "saa-plus",
  "level": "unknown",
  "reason": "unknown for its base model Segment Anything Model (SAM): training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "sam-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "sam",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "siglip-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "siglip",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "univad",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "vcp-clip-mvtec",
  "level": "test_seen",
  "reason": "in fine-tuning data: its test split was used for training",
  "sources": "https://arxiv.org/pdf/2407.12276v1 https://raw.githubusercontent.com/xiaozhen228/VCP-CLIP/main/utils/dataset.py"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "vcp-clip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad",
  "model": "winclip",
  "level": "unknown",
  "reason": "unknown for its base model OpenCLIP (LAION-400M checkpoints): training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "aa-clip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "aa-clip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "adaclip-all",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "adaclip-mvtec-colondb",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "adaclip-visa-clinicdb",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "adaptclip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "adaptclip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "anomaly-ov",
  "level": "unknown",
  "reason": "unknown for its base model LLaVA-OneVision: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "anomalyclip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "anomalyclip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "anomalydino",
  "level": "unknown",
  "reason": "unknown for its base model DINOv2: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "anomalygpt-mvtec",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "anomalygpt-supervised",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "anomalygpt-visa",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "april-gan-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "april-gan-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "clip-openai",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "dinov2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "dinov3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "filo-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "filo-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "imagenet-wrn50",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/pytorch/vision/main/torchvision/models/resnet.py"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "inctrl-mvtec",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.06495v3 https://raw.githubusercontent.com/mala-lab/InCTRL/main/open_clip/config/defaults.py https://laion.ai/blog/laion-400-open-dataset/ https://arxiv.org/pdf/2503.21622v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "inctrl-visa",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.06495v3 https://raw.githubusercontent.com/mala-lab/InCTRL/main/open_clip/config/defaults.py https://laion.ai/blog/laion-400-open-dataset/ https://arxiv.org/pdf/2503.21622v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "metauas",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "musc",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "openclip-laion2b",
  "level": "not_seen",
  "reason": "its test items were published from 2025-03-27, after the model's stated training-data cutoff (2022-01-31)",
  "sources": "https://arxiv.org/pdf/2210.08402v1 https://arxiv.org/pdf/2503.21622v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "openclip-laion400m",
  "level": "not_seen",
  "reason": "its test items were published from 2025-03-27, after the model's stated training-data cutoff (2021-12-31)",
  "sources": "https://laion.ai/blog/laion-400-open-dataset/ https://arxiv.org/pdf/2503.21622v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "saa-plus",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "sam-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "sam",
  "level": "not_seen",
  "reason": "its test items were published from 2025-03-27, after the model's stated training-data cutoff (2022-12-31)",
  "sources": "https://arxiv.org/pdf/2304.02643v1 https://arxiv.org/pdf/2503.21622v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "siglip-2",
  "level": "not_seen",
  "reason": "its test items were published from 2025-03-27, after the model's stated training-data cutoff (2022-12-31)",
  "sources": "https://arxiv.org/pdf/2209.06794v4 https://arxiv.org/pdf/2503.21622v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "siglip",
  "level": "not_seen",
  "reason": "its test items were published from 2025-03-27, after the model's stated training-data cutoff (2022-12-31)",
  "sources": "https://arxiv.org/pdf/2209.06794v4 https://arxiv.org/pdf/2503.21622v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "univad",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "vcp-clip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "vcp-clip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-ad-2",
  "model": "winclip",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2303.14814v1 https://laion.ai/blog/laion-400-open-dataset/ https://arxiv.org/pdf/2503.21622v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "aa-clip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "aa-clip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "adaclip-all",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "adaclip-mvtec-colondb",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "adaclip-visa-clinicdb",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "adaptclip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "adaptclip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "anomaly-ov",
  "level": "unknown",
  "reason": "unknown for its base model LLaVA-OneVision: unknown for its base model SigLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "anomalyclip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "anomalyclip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "anomalydino",
  "level": "unknown",
  "reason": "unknown for its base model DINOv2: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "anomalygpt-mvtec",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "anomalygpt-supervised",
  "level": "seen_split_unclear",
  "reason": "in fine-tuning data: the dataset was used; the source does not say which part",
  "sources": "https://raw.githubusercontent.com/CASIA-IVA-Lab/AnomalyGPT/main/README.md"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "anomalygpt-visa",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "april-gan-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "april-gan-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "clip-openai",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "dinov2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "dinov3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "filo-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "filo-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "imagenet-wrn50",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/pytorch/vision/main/torchvision/models/resnet.py"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "inctrl-mvtec",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.06495v3 https://raw.githubusercontent.com/mala-lab/InCTRL/main/open_clip/config/defaults.py https://laion.ai/blog/laion-400-open-dataset/ https://link.springer.com/article/10.1007/s11263-022-01578-9"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "inctrl-visa",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.06495v3 https://raw.githubusercontent.com/mala-lab/InCTRL/main/open_clip/config/defaults.py https://laion.ai/blog/laion-400-open-dataset/ https://link.springer.com/article/10.1007/s11263-022-01578-9"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "metauas",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "musc",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "openclip-laion2b",
  "level": "not_seen",
  "reason": "its test items were published from 2022-02-22, after the model's stated training-data cutoff (2022-01-31)",
  "sources": "https://arxiv.org/pdf/2210.08402v1 https://link.springer.com/article/10.1007/s11263-022-01578-9"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "openclip-laion400m",
  "level": "not_seen",
  "reason": "its test items were published from 2022-02-22, after the model's stated training-data cutoff (2021-12-31)",
  "sources": "https://laion.ai/blog/laion-400-open-dataset/ https://link.springer.com/article/10.1007/s11263-022-01578-9"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "saa-plus",
  "level": "unknown",
  "reason": "unknown for its base model Segment Anything Model (SAM): training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "sam-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "sam",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "siglip-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "siglip",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "univad",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "vcp-clip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "vcp-clip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mvtec-loco-ad",
  "model": "winclip",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2303.14814v1 https://laion.ai/blog/laion-400-open-dataset/ https://link.springer.com/article/10.1007/s11263-022-01578-9"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "aa-clip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "aa-clip-visa",
  "level": "seen_split_unclear",
  "reason": "in fine-tuning data: the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2503.06661v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "adaclip-all",
  "level": "test_seen",
  "reason": "in fine-tuning data: its test split was used for training",
  "sources": "https://raw.githubusercontent.com/caoyunkang/AdaCLIP/main/README.md https://raw.githubusercontent.com/caoyunkang/AdaCLIP/main/dataset/base_dataset.py"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "adaclip-mvtec-colondb",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "adaclip-visa-clinicdb",
  "level": "test_seen",
  "reason": "in fine-tuning data: its test split was used for training",
  "sources": "https://raw.githubusercontent.com/caoyunkang/AdaCLIP/main/README.md https://raw.githubusercontent.com/caoyunkang/AdaCLIP/main/dataset/base_dataset.py"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "adaptclip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "adaptclip-visa",
  "level": "test_seen",
  "reason": "in fine-tuning data: its test split was used for training",
  "sources": "https://arxiv.org/pdf/2505.09926v2"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "anomaly-ov",
  "level": "seen_split_unclear",
  "reason": "in fine-tuning data: the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2502.07601v2"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "anomalyclip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "anomalyclip-visa",
  "level": "test_seen",
  "reason": "in fine-tuning data: its test split was used for training",
  "sources": "https://arxiv.org/pdf/2310.18961v12 https://raw.githubusercontent.com/zqhang/AnomalyCLIP/main/train.sh"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "anomalydino",
  "level": "unknown",
  "reason": "unknown for its base model DINOv2: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "anomalygpt-mvtec",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "anomalygpt-supervised",
  "level": "seen_split_unclear",
  "reason": "in fine-tuning data: the dataset was used; the source does not say which part",
  "sources": "https://raw.githubusercontent.com/CASIA-IVA-Lab/AnomalyGPT/main/README.md"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "anomalygpt-visa",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2308.15366v4"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "april-gan-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "april-gan-visa",
  "level": "test_seen",
  "reason": "in fine-tuning data: its test split was used for training",
  "sources": "https://arxiv.org/pdf/2305.17382v3"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "clip-openai",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "dinov2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "dinov3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "filo-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "filo-visa",
  "level": "test_seen",
  "reason": "in fine-tuning data: its test split was used for training",
  "sources": "https://arxiv.org/pdf/2404.13671v3"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "imagenet-wrn50",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/pytorch/vision/main/torchvision/models/resnet.py"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "inctrl-mvtec",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.06495v3 https://raw.githubusercontent.com/mala-lab/InCTRL/main/open_clip/config/defaults.py https://laion.ai/blog/laion-400-open-dataset/ https://arxiv.org/pdf/2207.14315v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "inctrl-visa",
  "level": "test_seen",
  "reason": "in fine-tuning data: the whole dataset was used",
  "sources": "https://arxiv.org/pdf/2403.06495v3"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "metauas",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "musc",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "openclip-laion2b",
  "level": "not_seen",
  "reason": "its test items were published from 2022-07-28, after the model's stated training-data cutoff (2022-01-31)",
  "sources": "https://arxiv.org/pdf/2210.08402v1 https://arxiv.org/pdf/2207.14315v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "openclip-laion400m",
  "level": "not_seen",
  "reason": "its test items were published from 2022-07-28, after the model's stated training-data cutoff (2021-12-31)",
  "sources": "https://laion.ai/blog/laion-400-open-dataset/ https://arxiv.org/pdf/2207.14315v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "saa-plus",
  "level": "unknown",
  "reason": "unknown for its base model Segment Anything Model (SAM): training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "sam-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "sam",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "siglip-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "siglip",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "univad",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "vcp-clip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "vcp-clip-visa",
  "level": "test_seen",
  "reason": "in fine-tuning data: its test split was used for training",
  "sources": "https://arxiv.org/pdf/2407.12276v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "visa",
  "model": "winclip",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2303.14814v1 https://laion.ai/blog/laion-400-open-dataset/ https://arxiv.org/pdf/2207.14315v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "aa-clip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "aa-clip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "adaclip-all",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "adaclip-mvtec-colondb",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "adaclip-visa-clinicdb",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "adaptclip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "adaptclip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "anomaly-ov",
  "level": "unknown",
  "reason": "unknown for its base model LLaVA-OneVision: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "anomalyclip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "anomalyclip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "anomalydino",
  "level": "unknown",
  "reason": "unknown for its base model DINOv2: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "anomalygpt-mvtec",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "anomalygpt-supervised",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "anomalygpt-visa",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "april-gan-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "april-gan-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "clip-openai",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "dinov2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "dinov3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "filo-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "filo-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "imagenet-wrn50",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/pytorch/vision/main/torchvision/models/resnet.py"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "inctrl-mvtec",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.06495v3 https://raw.githubusercontent.com/mala-lab/InCTRL/main/open_clip/config/defaults.py https://laion.ai/blog/laion-400-open-dataset/ https://arxiv.org/pdf/2403.12580v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "inctrl-visa",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.06495v3 https://raw.githubusercontent.com/mala-lab/InCTRL/main/open_clip/config/defaults.py https://laion.ai/blog/laion-400-open-dataset/ https://arxiv.org/pdf/2403.12580v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "metauas",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "musc",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "openclip-laion2b",
  "level": "not_seen",
  "reason": "its test items were published from 2024-03-19, after the model's stated training-data cutoff (2022-01-31)",
  "sources": "https://arxiv.org/pdf/2210.08402v1 https://arxiv.org/pdf/2403.12580v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "openclip-laion400m",
  "level": "not_seen",
  "reason": "its test items were published from 2024-03-19, after the model's stated training-data cutoff (2021-12-31)",
  "sources": "https://laion.ai/blog/laion-400-open-dataset/ https://arxiv.org/pdf/2403.12580v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "saa-plus",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "sam-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "sam",
  "level": "not_seen",
  "reason": "its test items were published from 2024-03-19, after the model's stated training-data cutoff (2022-12-31)",
  "sources": "https://arxiv.org/pdf/2304.02643v1 https://arxiv.org/pdf/2403.12580v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "siglip-2",
  "level": "not_seen",
  "reason": "its test items were published from 2024-03-19, after the model's stated training-data cutoff (2022-12-31)",
  "sources": "https://arxiv.org/pdf/2209.06794v4 https://arxiv.org/pdf/2403.12580v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "siglip",
  "level": "not_seen",
  "reason": "its test items were published from 2024-03-19, after the model's stated training-data cutoff (2022-12-31)",
  "sources": "https://arxiv.org/pdf/2209.06794v4 https://arxiv.org/pdf/2403.12580v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "univad",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "vcp-clip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "vcp-clip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "real-iad",
  "model": "winclip",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2303.14814v1 https://laion.ai/blog/laion-400-open-dataset/ https://arxiv.org/pdf/2403.12580v1"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "aa-clip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "aa-clip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "adaclip-all",
  "level": "test_seen",
  "reason": "in fine-tuning data: its test split was used for training",
  "sources": "https://raw.githubusercontent.com/caoyunkang/AdaCLIP/main/README.md https://raw.githubusercontent.com/caoyunkang/AdaCLIP/main/dataset/base_dataset.py"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "adaclip-mvtec-colondb",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "adaclip-visa-clinicdb",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "adaptclip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "adaptclip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "anomaly-ov",
  "level": "unknown",
  "reason": "unknown for its base model LLaVA-OneVision: unknown for its base model SigLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "anomalyclip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "anomalyclip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "anomalydino",
  "level": "unknown",
  "reason": "unknown for its base model DINOv2: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "anomalygpt-mvtec",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "anomalygpt-supervised",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "anomalygpt-visa",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "april-gan-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "april-gan-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "clip-openai",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "dinov2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "dinov3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "filo-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "filo-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "imagenet-wrn50",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/pytorch/vision/main/torchvision/models/resnet.py"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "inctrl-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenCLIP (LAION-400M checkpoints): training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "inctrl-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenCLIP (LAION-400M checkpoints): training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "metauas",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "musc",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "openclip-laion2b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "openclip-laion400m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "saa-plus",
  "level": "unknown",
  "reason": "unknown for its base model Segment Anything Model (SAM): training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "sam-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "sam",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "siglip-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "siglip",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "univad",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "vcp-clip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "vcp-clip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "mpdd",
  "model": "winclip",
  "level": "unknown",
  "reason": "unknown for its base model OpenCLIP (LAION-400M checkpoints): training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "aa-clip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "aa-clip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "adaclip-all",
  "level": "test_seen",
  "reason": "in fine-tuning data: its test split was used for training",
  "sources": "https://raw.githubusercontent.com/caoyunkang/AdaCLIP/main/README.md https://raw.githubusercontent.com/caoyunkang/AdaCLIP/main/dataset/base_dataset.py"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "adaclip-mvtec-colondb",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "adaclip-visa-clinicdb",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "adaptclip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "adaptclip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "anomaly-ov",
  "level": "unknown",
  "reason": "unknown for its base model LLaVA-OneVision: unknown for its base model SigLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "anomalyclip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "anomalyclip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "anomalydino",
  "level": "unknown",
  "reason": "unknown for its base model DINOv2: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "anomalygpt-mvtec",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "anomalygpt-supervised",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "anomalygpt-visa",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "april-gan-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "april-gan-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "clip-openai",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "dinov2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "dinov3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "filo-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "filo-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "imagenet-wrn50",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/pytorch/vision/main/torchvision/models/resnet.py"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "inctrl-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenCLIP (LAION-400M checkpoints): training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "inctrl-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenCLIP (LAION-400M checkpoints): training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "metauas",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "musc",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "openclip-laion2b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "openclip-laion400m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "saa-plus",
  "level": "unknown",
  "reason": "unknown for its base model Segment Anything Model (SAM): training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "sam-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "sam",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "siglip-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "siglip",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "univad",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "vcp-clip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "vcp-clip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "btad",
  "model": "winclip",
  "level": "unknown",
  "reason": "unknown for its base model OpenCLIP (LAION-400M checkpoints): training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "aa-clip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "aa-clip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "adaclip-all",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "adaclip-mvtec-colondb",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "adaclip-visa-clinicdb",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "adaptclip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "adaptclip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "anomaly-ov",
  "level": "unknown",
  "reason": "unknown for its base model LLaVA-OneVision: unknown for its base model SigLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "anomalyclip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "anomalyclip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "anomalydino",
  "level": "unknown",
  "reason": "unknown for its base model DINOv2: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "anomalygpt-mvtec",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "anomalygpt-supervised",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "anomalygpt-visa",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "april-gan-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "april-gan-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "clip-openai",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "dinov2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "dinov3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "filo-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "filo-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "imagenet-wrn50",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/pytorch/vision/main/torchvision/models/resnet.py"
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "inctrl-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenCLIP (LAION-400M checkpoints): training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "inctrl-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenCLIP (LAION-400M checkpoints): training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "metauas",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "musc",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "openclip-laion2b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "openclip-laion400m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "saa-plus",
  "level": "unknown",
  "reason": "unknown for its base model Segment Anything Model (SAM): training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "sam-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "sam",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "siglip-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "siglip",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "univad",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "vcp-clip-mvtec",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "vcp-clip-visa",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "industrial-ad",
  "dataset": "ksdd2",
  "model": "winclip",
  "level": "unknown",
  "reason": "unknown for its base model OpenCLIP (LAION-400M checkpoints): training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "amber",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "bloom",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "claude-mythos-preview",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "claude-opus-4-5",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://assets.anthropic.com/m/64823ba7485345a7/Claude-Opus-4-5-System-Card.pdf"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "dclm-7b-it",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (DCLM's MMLU overlap rule (question last sentence + one answer option on the same page; recall-oriented, likely overestimate) flagged 0.007% of DCLM-Baseline documents (vs 0.001% Dolma-V1.7, 0.009% FineWeb-Edu). Removing the flagged overlaps did not lower MMLU accuracy at 7B-2x scale (51.8 -> 52.7).) (via DCLM-Baseline 1.0)",
  "sources": "https://huggingface.co/mlfoundations/dclm-7b-it/raw/main/README.md https://arxiv.org/pdf/2406.11794v4"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "dclm-7b",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (DCLM's MMLU overlap rule (question last sentence + one answer option on the same page; recall-oriented, likely overestimate) flagged 0.007% of DCLM-Baseline documents (vs 0.001% Dolma-V1.7, 0.009% FineWeb-Edu). Removing the flagged overlaps did not lower MMLU accuracy at 7B-2x scale (51.8 -> 52.7).) (via DCLM-Baseline 1.0)",
  "sources": "https://huggingface.co/apple/DCLM-7B/raw/main/README.md https://arxiv.org/pdf/2406.11794v4"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "deepseek-r1-0528",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "deepseek-r1-distill",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "deepseek-r1",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2501.12948v2"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "deepseek-v3-0324",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "deepseek-v3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "deepseek-v3.2-exp",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "deepseek-v3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "deepseek-v3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "deepseek-v4.1-flash",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "deepseek-v4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "gemma-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "gpt-4",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (~0.6% of 1000 sampled MMLU test examples matched GPT-4 pre-training data (50-character substring match))",
  "sources": "https://arxiv.org/pdf/2303.08774v6"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "gpt-6-astra",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "gpt-oss",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "internvl3.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "internvl3",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "k2-65b",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (MMLU test questions are also substantially 'dirty' in Pile-train: DCLM-baseline (27.70%) has only 1.1x more dirty entries than Pile-train. Exact Pile-train rate is in Table 2 (not reliably extracted from the PDF text).) [third-party study] (via The Pile)",
  "sources": "https://huggingface.co/datasets/IFM/K2Datasets/raw/main/README.md https://arxiv.org/pdf/2506.12229v4"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "k2-horizon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "k2-v2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "llama-2",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (MMLU-Humanities (L=50): 709 of 4,705 samples (15%) 'not clean' (>=20% token contamination) and 520 (11%) 'dirty' (>=80%); flagged as boosted by contamination. MMLU-Overall for 70B: 2,180 of 14,042 (15.5%) not clean, 1,536 (10.9%) dirty, with a small effect on the overall score (-0.9 clean-vs-mean delta).)",
  "sources": "https://arxiv.org/pdf/2307.09288v2 https://arxiv.org/html/2307.09288v2"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "llama-3.1",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (8-gram overlap gave contamination scores too high to estimate a performance gain, even at higher thresholds; percentage not reported (dash in Table 15).)",
  "sources": "https://arxiv.org/pdf/2407.21783v3"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "llama-3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "llama-3.3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "llama-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "llava-onevision-1.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "llava-onevision",
  "level": "unknown",
  "reason": "unknown for its base model SigLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "ministral-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "mistral-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "mixtral-8x7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "molmo",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "molmo2",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "olmo-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "olmo-2-1124",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (DCLM's MMLU overlap rule (question last sentence + one answer option on the same page; recall-oriented, likely overestimate) flagged 0.007% of DCLM-Baseline documents (vs 0.001% Dolma-V1.7, 0.009% FineWeb-Edu). Removing the flagged overlaps did not lower MMLU accuracy at 7B-2x scale (51.8 -> 52.7).) (via DCLM-Baseline 1.0)",
  "sources": "https://huggingface.co/datasets/allenai/olmo-mix-1124/raw/main/README.md https://arxiv.org/pdf/2406.11794v4"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "olmo-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "olmo-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "olmo-hybrid-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "phi-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "phi-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "phi-4-mini-flash-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "phi-4-mini-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "phi-4-mini",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "phi-4-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "phi-4",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2412.08905v1"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "phi-slimmoe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "pythia-deduped",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (MMLU test questions are also substantially 'dirty' in Pile-train: DCLM-baseline (27.70%) has only 1.1x more dirty entries than Pile-train. Exact Pile-train rate is in Table 2 (not reliably extracted from the PDF text).) [third-party study] (via The Pile)",
  "sources": "https://huggingface.co/EleutherAI/pythia-12b-deduped/raw/main/README.md https://huggingface.co/EleutherAI/pythia-14m-deduped/raw/main/README.md https://huggingface.co/EleutherAI/pythia-31m-deduped/raw/main/README.md https://arxiv.org/pdf/2506.12229v4"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "pythia",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (MMLU test questions are also substantially 'dirty' in Pile-train: DCLM-baseline (27.70%) has only 1.1x more dirty entries than Pile-train. Exact Pile-train rate is in Table 2 (not reliably extracted from the PDF text).) [third-party study] (via The Pile)",
  "sources": "https://huggingface.co/EleutherAI/pythia-12b/raw/main/README.md https://huggingface.co/EleutherAI/pythia-14m/raw/main/README.md https://huggingface.co/EleutherAI/pythia-31m/raw/main/README.md https://arxiv.org/pdf/2506.12229v4"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "qwen2.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "qwen3-2507",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "qwen3-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "qwen3.6",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "qwen3.8-flash-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "qwen3.8",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "qwen3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "smollm2",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (DCLM's MMLU overlap rule (question last sentence + one answer option on the same page; recall-oriented, likely overestimate) flagged 0.007% of DCLM-Baseline documents (vs 0.001% Dolma-V1.7, 0.009% FineWeb-Edu). Removing the flagged overlaps did not lower MMLU accuracy at 7B-2x scale (51.8 -> 52.7).) (via DCLM-Baseline 1.0); this contradicts a statement that the data was excluded",
  "sources": "https://arxiv.org/pdf/2502.02737v1 https://arxiv.org/pdf/2406.11794v4"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "mmlu",
  "model": "smollm3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "amber",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "bloom",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "claude-mythos-preview",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "dclm-7b-it",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "dclm-7b",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (Via ProofPile2: Llemma Table 6 lists 30-gram hits between GSM8k test problems and OpenWebMath documents (2 test examples, 3 documents; 0 for solutions); CONDA 2024 report lists gsm8k as contaminated in EleutherAI/proof-pile-2.) [third-party study]",
  "sources": "https://arxiv.org/pdf/2310.10631v3 https://arxiv.org/pdf/2407.21530v2"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "deepseek-r1-0528",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "deepseek-r1-distill",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "deepseek-r1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "deepseek-v3-0324",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "deepseek-v3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "deepseek-v3.2-exp",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "deepseek-v3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "deepseek-v3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "deepseek-v4.1-flash",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "deepseek-v4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "gemma-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "gpt-4",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (~1% of 1000 sampled GSM-8K test examples matched GPT-4 pre-training data)",
  "sources": "https://arxiv.org/pdf/2303.08774v6"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "gpt-6-astra",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "gpt-oss",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "internvl3.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "internvl3",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "k2-65b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "k2-horizon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "k2-v2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "llama-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "llama-3.1",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (41% of GSM8K examples considered contaminated (8-gram overlap); estimated performance gain 0.0 / 0.1 / 1.3 points for 8B / 70B / 405B.)",
  "sources": "https://arxiv.org/html/2407.21783v3"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "llama-3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "llama-3.3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "llama-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "llava-onevision-1.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "llava-onevision",
  "level": "unknown",
  "reason": "unknown for its base model SigLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "ministral-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "mistral-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "mixtral-8x7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "molmo",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "molmo2",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "olmo-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "olmo-2-1124",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2501.00656v3 https://arxiv.org/pdf/2411.15124v5"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "olmo-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "olmo-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "olmo-hybrid-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "phi-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "phi-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "phi-4-mini-flash-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "phi-4-mini-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "phi-4-mini",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "phi-4-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "phi-4",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2412.08905v1"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "phi-slimmoe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "pythia-deduped",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "pythia",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "qwen2.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "qwen3-2507",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "qwen3-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "qwen3.6",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "qwen3.8-flash-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "qwen3.8",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "qwen3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "smollm2",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2502.02737v1"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "gsm8k",
  "model": "smollm3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "amber",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "bloom",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "claude-mythos-preview",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "dclm-7b-it",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "dclm-7b",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (Via ProofPile2 (Llemma paper): ~7% of MATH test problem statements and 0.6% of MATH test solutions have 30-gram hits in Proof-Pile-2 (OpenWebMath/AlgebraicStack). CONDA 2024 report also lists hendrycks/competition_math and gsm8k as contaminated in EleutherAI/proof-pile-2.) [third-party study]",
  "sources": "https://arxiv.org/pdf/2310.10631v3 https://arxiv.org/pdf/2407.21530v2"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "deepseek-r1-0528",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "deepseek-r1-distill",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "deepseek-r1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "deepseek-v3-0324",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "deepseek-v3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "deepseek-v3.2-exp",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "deepseek-v3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "deepseek-v3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "deepseek-v4.1-flash",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "deepseek-v4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "gemma-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "gpt-4",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2303.08774v6"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "gpt-6-astra",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "gpt-oss",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "internvl3.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "internvl3",
  "level": "related",
  "reason": "via base model Qwen2.5: a contamination analysis found test examples of the related dataset MATH-500 (Exact completion of 21.2% of MATH-500 problems from a 60% prefix (Table 2).) [third-party study]",
  "sources": "https://arxiv.org/pdf/2504.10479v3 https://arxiv.org/pdf/2507.10532v3 https://raw.githubusercontent.com/openai/prm800k/main/README.md https://huggingface.co/datasets/HuggingFaceH4/MATH-500/raw/main/README.md"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "k2-65b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "k2-horizon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "k2-v2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "llama-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "llama-3.1",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (1% of MATH examples considered contaminated at the selected threshold; estimated gain 0.0 / -0.1 / -0.2. Authors note low thresholds yield high contamination for MATH but no performance gains.)",
  "sources": "https://arxiv.org/html/2407.21783v3"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "llama-3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "llama-3.3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "llama-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "llava-onevision-1.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "llava-onevision",
  "level": "unknown",
  "reason": "unknown for its base model SigLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "ministral-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "mistral-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "mixtral-8x7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "molmo",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "molmo2",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "olmo-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "olmo-2-1124",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2501.00656v3 https://arxiv.org/pdf/2411.15124v5"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "olmo-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "olmo-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "olmo-hybrid-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "phi-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "phi-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "phi-4-mini-flash-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "phi-4-mini-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "phi-4-mini",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "phi-4-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "phi-4",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2412.08905v1"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "phi-slimmoe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "pythia-deduped",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "pythia",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "qwen2.5-vl",
  "level": "related",
  "reason": "via base model Qwen2.5: a contamination analysis found test examples of the related dataset MATH-500 (Exact completion of 21.2% of MATH-500 problems from a 60% prefix (Table 2).) [third-party study]",
  "sources": "https://arxiv.org/pdf/2502.13923v1 https://arxiv.org/pdf/2507.10532v3 https://raw.githubusercontent.com/openai/prm800k/main/README.md https://huggingface.co/datasets/HuggingFaceH4/MATH-500/raw/main/README.md"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "qwen2.5",
  "level": "related",
  "reason": "a contamination analysis found test examples of the related dataset MATH-500 (Exact completion of 21.2% of MATH-500 problems from a 60% prefix (Table 2).) [third-party study]",
  "sources": "https://arxiv.org/pdf/2507.10532v3 https://raw.githubusercontent.com/openai/prm800k/main/README.md https://huggingface.co/datasets/HuggingFaceH4/MATH-500/raw/main/README.md"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "qwen3-2507",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "qwen3-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "qwen3.6",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "qwen3.8-flash-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "qwen3.8",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "qwen3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "smollm2",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data (via SmolTalk)",
  "sources": "https://arxiv.org/pdf/2502.02737v1 https://huggingface.co/HuggingFaceTB/SmolLM2-1.7B-Instruct/raw/main/README.md https://huggingface.co/datasets/HuggingFaceTB/smoltalk/raw/main/README.md https://huggingface.co/datasets/meta-math/MetaMathQA/raw/main/README.md"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "math",
  "model": "smollm3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "amber",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "bloom",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "claude-mythos-preview",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "dclm-7b-it",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "dclm-7b",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (Via StarCoder data: Riddell et al. (2024) find 18.9% of HumanEval gold solutions (and 20.8% of MBPP) seen in the Python split of the Stack training data of StarCoderBase (60.40 GB searched), despite the Stack's string-matching decontamination.) [third-party study]",
  "sources": "https://arxiv.org/pdf/2403.04811v1"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "deepseek-r1-0528",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "deepseek-r1-distill",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "deepseek-r1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "deepseek-v3-0324",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "deepseek-v3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "deepseek-v3.2-exp",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "deepseek-v3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "deepseek-v3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "deepseek-v4.1-flash",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "deepseek-v4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "gemma-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "gpt-4",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (25% of HumanEval problems matched GPT-4 pre-training data (whole benchmark checked); score on the non-contaminated subset was 65.58% vs 67.0% overall)",
  "sources": "https://arxiv.org/pdf/2303.08774v6"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "gpt-6-astra",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "gpt-oss",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "internvl3.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "internvl3",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "k2-65b",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (Riddell et al. (2024): 12.2% of HumanEval gold solutions have a perfect (surface- or semantic-level) match in the Pile's GitHub split; CONDA 2024 contamination report also lists openai_humaneval as found (>0%) in EleutherAI/pile. (WIMBD's exact-match input-field check reports 0.00% for openai-humaneval in the Pile, Table 16.)) [third-party study] (via The Pile)",
  "sources": "https://huggingface.co/datasets/IFM/K2Datasets/raw/main/README.md https://arxiv.org/pdf/2403.04811v1 https://arxiv.org/pdf/2407.21530v2"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "k2-horizon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "k2-v2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "llama-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "llama-3.1",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (8-gram overlap gave contamination scores too high to estimate a performance gain, even at higher thresholds; percentage not reported (dash in Table 15).)",
  "sources": "https://arxiv.org/pdf/2407.21783v3"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "llama-3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "llama-3.3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "llama-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "llava-onevision-1.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "llava-onevision",
  "level": "unknown",
  "reason": "unknown for its base model SigLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "ministral-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "mistral-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "mixtral-8x7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "molmo",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "molmo2",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "olmo-1",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (WIMBD analysis by the Dolma authors: over 90% of the OpenAI HumanEval test set appears in Dolma documents (mostly via the GitHub code subset); only Paloma was decontaminated, not downstream benchmarks.) (via Dolma v1.5)",
  "sources": "https://huggingface.co/datasets/allenai/dolma/raw/main/README.md https://arxiv.org/pdf/2402.00159v2"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "olmo-2-1124",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "olmo-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "olmo-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "olmo-hybrid-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "phi-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "phi-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "phi-4-mini-flash-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "phi-4-mini-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "phi-4-mini",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "phi-4-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "phi-4",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2412.08905v1"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "phi-slimmoe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "pythia-deduped",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (Riddell et al. (2024): 12.2% of HumanEval gold solutions have a perfect (surface- or semantic-level) match in the Pile's GitHub split; CONDA 2024 contamination report also lists openai_humaneval as found (>0%) in EleutherAI/pile. (WIMBD's exact-match input-field check reports 0.00% for openai-humaneval in the Pile, Table 16.)) [third-party study] (via The Pile)",
  "sources": "https://huggingface.co/EleutherAI/pythia-12b-deduped/raw/main/README.md https://huggingface.co/EleutherAI/pythia-14m-deduped/raw/main/README.md https://huggingface.co/EleutherAI/pythia-31m-deduped/raw/main/README.md https://arxiv.org/pdf/2403.04811v1 https://arxiv.org/pdf/2407.21530v2"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "pythia",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (Riddell et al. (2024): 12.2% of HumanEval gold solutions have a perfect (surface- or semantic-level) match in the Pile's GitHub split; CONDA 2024 contamination report also lists openai_humaneval as found (>0%) in EleutherAI/pile. (WIMBD's exact-match input-field check reports 0.00% for openai-humaneval in the Pile, Table 16.)) [third-party study] (via The Pile)",
  "sources": "https://huggingface.co/EleutherAI/pythia-12b/raw/main/README.md https://huggingface.co/EleutherAI/pythia-14m/raw/main/README.md https://huggingface.co/EleutherAI/pythia-31m/raw/main/README.md https://arxiv.org/pdf/2403.04811v1 https://arxiv.org/pdf/2407.21530v2"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "qwen2.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "qwen3-2507",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "qwen3-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "qwen3.6",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "qwen3.8-flash-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "qwen3.8",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "qwen3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "smollm2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "humaneval",
  "model": "smollm3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "amber",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "bloom",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "claude-mythos-preview",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "dclm-7b-it",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (32.6% of ARC-Challenge test questions are 'dirty' in DCLM-baseline (17x more than in Pile-train).) [third-party study] (via DCLM-Baseline 1.0)",
  "sources": "https://huggingface.co/mlfoundations/dclm-7b-it/raw/main/README.md https://arxiv.org/pdf/2506.12229v4"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "dclm-7b",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (32.6% of ARC-Challenge test questions are 'dirty' in DCLM-baseline (17x more than in Pile-train).) [third-party study] (via DCLM-Baseline 1.0)",
  "sources": "https://huggingface.co/apple/DCLM-7B/raw/main/README.md https://arxiv.org/pdf/2406.11794v4 https://arxiv.org/pdf/2506.12229v4"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "deepseek-r1-0528",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "deepseek-r1-distill",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "deepseek-r1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "deepseek-v3-0324",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "deepseek-v3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "deepseek-v3.2-exp",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "deepseek-v3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "deepseek-v3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "deepseek-v4.1-flash",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "deepseek-v4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "gemma-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "gpt-4",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (~3.4% of 1000 sampled ARC (Challenge set) test examples matched GPT-4 pre-training data)",
  "sources": "https://arxiv.org/pdf/2303.08774v6"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "gpt-6-astra",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "gpt-oss",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "internvl3.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "internvl3",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "k2-65b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "k2-horizon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "k2-v2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "llama-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "llama-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "llama-3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "llama-3.3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "llama-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "llava-onevision-1.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "llava-onevision",
  "level": "unknown",
  "reason": "unknown for its base model SigLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "ministral-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "mistral-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "mixtral-8x7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "molmo",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "molmo2",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "olmo-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "olmo-2-1124",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (32.6% of ARC-Challenge test questions are 'dirty' in DCLM-baseline (17x more than in Pile-train).) [third-party study] (via DCLM-Baseline 1.0)",
  "sources": "https://huggingface.co/datasets/allenai/olmo-mix-1124/raw/main/README.md https://arxiv.org/pdf/2506.12229v4"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "olmo-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "olmo-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "olmo-hybrid-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "phi-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "phi-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "phi-4-mini-flash-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "phi-4-mini-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "phi-4-mini",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "phi-4-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "phi-4",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2412.08905v1"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "phi-slimmoe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "pythia-deduped",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "pythia",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "qwen2.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "qwen3-2507",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "qwen3-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "qwen3.6",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "qwen3.8-flash-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "qwen3.8",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "qwen3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "smollm2",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (32.6% of ARC-Challenge test questions are 'dirty' in DCLM-baseline (17x more than in Pile-train).) [third-party study] (via DCLM-Baseline 1.0)",
  "sources": "https://arxiv.org/pdf/2502.02737v1 https://arxiv.org/pdf/2506.12229v4"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "arc-challenge",
  "model": "smollm3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "amber",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "bloom",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "claude-mythos-preview",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "dclm-7b-it",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (Overlaps between HellaSwag test examples and DCLM-Baseline pages were detected and removed for an ablation (fraction not reported in text); removal did not lower HellaSwag accuracy (77.9 -> 78.4). In the 10-gram token-overlap analysis HellaSwag is the notable exception with a performance gap between full and 'clean' subsets (Figure 12).) (via DCLM-Baseline 1.0)",
  "sources": "https://huggingface.co/mlfoundations/dclm-7b-it/raw/main/README.md https://arxiv.org/pdf/2406.11794v4"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "dclm-7b",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (Overlaps between HellaSwag test examples and DCLM-Baseline pages were detected and removed for an ablation (fraction not reported in text); removal did not lower HellaSwag accuracy (77.9 -> 78.4). In the 10-gram token-overlap analysis HellaSwag is the notable exception with a performance gap between full and 'clean' subsets (Figure 12).) (via DCLM-Baseline 1.0)",
  "sources": "https://huggingface.co/apple/DCLM-7B/raw/main/README.md https://arxiv.org/pdf/2406.11794v4"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "deepseek-r1-0528",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "deepseek-r1-distill",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "deepseek-r1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "deepseek-v3-0324",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "deepseek-v3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "deepseek-v3.2-exp",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "deepseek-v3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "deepseek-v3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "deepseek-v4.1-flash",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "deepseek-v4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "gemma-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "gpt-4",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2303.08774v6"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "gpt-6-astra",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "gpt-oss",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "internvl3.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "internvl3",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "k2-65b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "k2-horizon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "k2-v2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "llama-2",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (With min. match length L=40: 2,651 of 10,042 samples (26%) 'not clean' (>=20% of tokens in 10+-token n-grams found in pre-training data) and 848 (8.4%) 'dirty' (>=80%); dirty samples scored 92.2 vs 82.5 sampling mean for 70B (Z=7.42). Authors conclude HellaSwag scores appear boosted by contamination, 70B more than 7B.)",
  "sources": "https://arxiv.org/pdf/2307.09288v2 https://arxiv.org/html/2307.09288v2"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "llama-3.1",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (85% of HellaSwag examples considered contaminated (8-gram overlap with the pre-training corpus at a dataset-specific token-ratio threshold); estimated performance gain 14.8 / 14.8 / 14.3 points for 8B / 70B / 405B.)",
  "sources": "https://arxiv.org/html/2407.21783v3 https://arxiv.org/pdf/2407.21783v3"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "llama-3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "llama-3.3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "llama-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "llava-onevision-1.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "llava-onevision",
  "level": "unknown",
  "reason": "unknown for its base model SigLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "ministral-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "mistral-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "mixtral-8x7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "molmo",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "molmo2",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "olmo-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "olmo-2-1124",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (Overlaps between HellaSwag test examples and DCLM-Baseline pages were detected and removed for an ablation (fraction not reported in text); removal did not lower HellaSwag accuracy (77.9 -> 78.4). In the 10-gram token-overlap analysis HellaSwag is the notable exception with a performance gap between full and 'clean' subsets (Figure 12).) (via DCLM-Baseline 1.0)",
  "sources": "https://huggingface.co/datasets/allenai/olmo-mix-1124/raw/main/README.md https://arxiv.org/pdf/2406.11794v4"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "olmo-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "olmo-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "olmo-hybrid-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "phi-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "phi-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "phi-4-mini-flash-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "phi-4-mini-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "phi-4-mini",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "phi-4-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "phi-4",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2412.08905v1"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "phi-slimmoe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "pythia-deduped",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "pythia",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "qwen2.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "qwen3-2507",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "qwen3-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "qwen3.6",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "qwen3.8-flash-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "qwen3.8",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "qwen3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "smollm2",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (Overlaps between HellaSwag test examples and DCLM-Baseline pages were detected and removed for an ablation (fraction not reported in text); removal did not lower HellaSwag accuracy (77.9 -> 78.4). In the 10-gram token-overlap analysis HellaSwag is the notable exception with a performance gap between full and 'clean' subsets (Figure 12).) (via DCLM-Baseline 1.0)",
  "sources": "https://arxiv.org/pdf/2502.02737v1 https://arxiv.org/pdf/2406.11794v4"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "hellaswag",
  "model": "smollm3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "amber",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "bloom",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "claude-mythos-preview",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "dclm-7b-it",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "dclm-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "deepseek-r1-0528",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "deepseek-r1-distill",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "deepseek-r1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "deepseek-v3-0324",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "deepseek-v3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "deepseek-v3.2-exp",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "deepseek-v3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "deepseek-v3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "deepseek-v4.1-flash",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "deepseek-v4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "gemma-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "gpt-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "gpt-6-astra",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "gpt-oss",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "internvl3.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "internvl3",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "k2-65b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "k2-horizon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "k2-v2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "llama-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "llama-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "llama-3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "llama-3.3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "llama-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "llava-onevision-1.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "llava-onevision",
  "level": "unknown",
  "reason": "unknown for its base model SigLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "ministral-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "mistral-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "mixtral-8x7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "molmo",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "molmo2",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "olmo-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "olmo-2-1124",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "olmo-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "olmo-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "olmo-hybrid-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "phi-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "phi-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "phi-4-mini-flash-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "phi-4-mini-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "phi-4-mini",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "phi-4-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "phi-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "phi-slimmoe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "pythia-deduped",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "pythia",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "qwen2.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "qwen3-2507",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "qwen3-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "qwen3.6",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "qwen3.8-flash-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "qwen3.8",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "qwen3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "smollm2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "truthfulqa",
  "model": "smollm3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "amber",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "bloom",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "claude-mythos-preview",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "dclm-7b-it",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "dclm-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "deepseek-r1-0528",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "deepseek-r1-distill",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "deepseek-r1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "deepseek-v3-0324",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "deepseek-v3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "deepseek-v3.2-exp",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "deepseek-v3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "deepseek-v3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "deepseek-v4.1-flash",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "deepseek-v4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "gemma-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "gpt-4",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (~0.9% of 1000 sampled WinoGrande test examples matched GPT-4 pre-training data)",
  "sources": "https://arxiv.org/pdf/2303.08774v6"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "gpt-6-astra",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "gpt-oss",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "internvl3.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "internvl3",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "k2-65b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "k2-horizon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "k2-v2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "llama-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "llama-3.1",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (6% of Winogrande examples considered contaminated; estimated gain -0.1 / -0.1 / -0.2 (no benefit).)",
  "sources": "https://arxiv.org/html/2407.21783v3"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "llama-3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "llama-3.3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "llama-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "llava-onevision-1.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "llava-onevision",
  "level": "unknown",
  "reason": "unknown for its base model SigLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "ministral-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "mistral-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "mixtral-8x7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "molmo",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "molmo2",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "olmo-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "olmo-2-1124",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "olmo-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "olmo-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "olmo-hybrid-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "phi-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "phi-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "phi-4-mini-flash-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "phi-4-mini-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "phi-4-mini",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "phi-4-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "phi-4",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2412.08905v1"
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "phi-slimmoe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "pythia-deduped",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "pythia",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "qwen2.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "qwen3-2507",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "qwen3-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "qwen3.6",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "qwen3.8-flash-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "qwen3.8",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "qwen3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "smollm2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-classic",
  "dataset": "winogrande",
  "model": "smollm3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "claude-mythos-preview",
  "level": "related",
  "reason": "a contamination analysis found test examples of the related dataset MMMU-Pro (A large fraction of MMMU-Pro images were found in the training data, mainly via textbooks, homework-help sites and document crawls. The developer omitted MMMU-Pro results from the system card.)",
  "sources": "https://www-cdn.anthropic.com/08ab9158070959f88f296514c21b7facce6f52bc.pdf https://arxiv.org/pdf/2409.02813v3 https://huggingface.co/datasets/MMMU/MMMU_Pro/raw/main/README.md"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "gpt-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "internvl3.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "internvl3",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "llava-onevision-1.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "llava-onevision",
  "level": "unknown",
  "reason": "unknown for its base model SigLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "molmo",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "molmo2",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "claude-mythos-preview",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the training data (A large fraction of MMMU-Pro images were found in the training data, mainly via textbooks, homework-help sites and document crawls. The developer omitted MMMU-Pro results from the system card.)",
  "sources": "https://www-cdn.anthropic.com/08ab9158070959f88f296514c21b7facce6f52bc.pdf"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "gpt-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "internvl3.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "internvl3",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "llava-onevision-1.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "llava-onevision",
  "level": "unknown",
  "reason": "unknown for its base model SigLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "molmo",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "molmo2",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mmmu-pro",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "claude-mythos-preview",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "gpt-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "internvl3.5",
  "level": "related",
  "reason": "in fine-tuning data: related dataset ChartQA was used",
  "sources": "https://arxiv.org/pdf/2508.18265v2 https://arxiv.org/pdf/2504.10479v3 https://arxiv.org/pdf/2412.05271v5 https://arxiv.org/pdf/2310.02255v3 https://huggingface.co/datasets/AI4Math/MathVista/raw/main/source.json"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "internvl3",
  "level": "related",
  "reason": "related dataset ChartQA was used",
  "sources": "https://arxiv.org/pdf/2504.10479v3 https://arxiv.org/pdf/2412.05271v5 https://arxiv.org/pdf/2310.02255v3 https://huggingface.co/datasets/AI4Math/MathVista/raw/main/source.json"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "llava-onevision-1.5",
  "level": "related",
  "reason": "in fine-tuning data: related dataset DocVQA was used",
  "sources": "https://huggingface.co/datasets/mvp-lab/LLaVA-OneVision-1.5-Instruct-Data/raw/main/README.md https://arxiv.org/pdf/2310.02255v3 https://huggingface.co/datasets/AI4Math/MathVista/raw/main/source.json"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "llava-onevision",
  "level": "related",
  "reason": "in fine-tuning data: related dataset ChartQA was used",
  "sources": "https://arxiv.org/pdf/2408.03326v3 https://arxiv.org/pdf/2310.02255v3 https://huggingface.co/datasets/AI4Math/MathVista/raw/main/source.json"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "molmo",
  "level": "related",
  "reason": "in fine-tuning data: related dataset ChartQA was used",
  "sources": "https://arxiv.org/pdf/2409.17146v2 https://arxiv.org/pdf/2310.02255v3 https://huggingface.co/datasets/AI4Math/MathVista/raw/main/source.json"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "molmo2",
  "level": "related",
  "reason": "in fine-tuning data: related dataset ChartQA was used",
  "sources": "https://arxiv.org/pdf/2601.10611v4 https://raw.githubusercontent.com/allenai/molmo2/main/launch_scripts/sft.py https://arxiv.org/pdf/2310.02255v3 https://huggingface.co/datasets/AI4Math/MathVista/raw/main/source.json"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "mathvista",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "claude-mythos-preview",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "gpt-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "internvl3.5",
  "level": "seen_split_unclear",
  "reason": "in fine-tuning data: the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2508.18265v2 https://arxiv.org/pdf/2504.10479v3 https://arxiv.org/pdf/2412.05271v5"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "internvl3",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2504.10479v3 https://arxiv.org/pdf/2412.05271v5"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "llava-onevision-1.5",
  "level": "seen_split_unclear",
  "reason": "in fine-tuning data: the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/datasets/mvp-lab/LLaVA-OneVision-1.5-Instruct-Data/raw/main/README.md"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "llava-onevision",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2408.03326v3"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "molmo",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2409.17146v2"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "molmo2",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2601.10611v4 https://raw.githubusercontent.com/allenai/molmo2/main/launch_scripts/sft.py"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "chartqa",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "claude-mythos-preview",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "gpt-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "internvl3.5",
  "level": "seen_split_unclear",
  "reason": "in fine-tuning data: the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2508.18265v2 https://arxiv.org/pdf/2504.10479v3 https://arxiv.org/pdf/2412.05271v5"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "internvl3",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2504.10479v3 https://arxiv.org/pdf/2412.05271v5"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "llava-onevision-1.5",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data",
  "sources": "https://huggingface.co/datasets/mvp-lab/LLaVA-OneVision-1.5-Instruct-Data/raw/main/README.md"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "llava-onevision",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2408.03326v3"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "molmo",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2409.17146v2"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "molmo2",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2601.10611v4 https://raw.githubusercontent.com/allenai/molmo2/main/launch_scripts/sft.py"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "docvqa",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "claude-mythos-preview",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "gpt-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "internvl3.5",
  "level": "seen_split_unclear",
  "reason": "in fine-tuning data: the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2508.18265v2 https://arxiv.org/pdf/2504.10479v3 https://arxiv.org/pdf/2412.05271v5"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "internvl3",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2504.10479v3 https://arxiv.org/pdf/2412.05271v5"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "llava-onevision-1.5",
  "level": "seen_split_unclear",
  "reason": "in fine-tuning data: the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/datasets/mvp-lab/LLaVA-OneVision-1.5-Instruct-Data/raw/main/README.md https://arxiv.org/pdf/2509.23661v3"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "llava-onevision",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2408.03326v3"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "molmo",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2409.17146v2"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "molmo2",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2601.10611v4 https://raw.githubusercontent.com/allenai/molmo2/main/launch_scripts/sft.py"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "ai2d",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "claude-3-5-sonnet",
  "level": "not_seen",
  "reason": "its test items were published from 2024-06-26, after the model's stated training-data cutoff (2024-04-01)",
  "sources": "https://web.archive.org/web/20241031205657id_/https://docs.anthropic.com/en/docs/about-claude/models https://arxiv.org/pdf/2406.18521v1"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "claude-3",
  "level": "not_seen",
  "reason": "its test items were published from 2024-06-26, after the model's stated training-data cutoff (2023-08-01)",
  "sources": "https://www-cdn.anthropic.com/de8ba9b01c9ab7cbabf5c33b80b7bbc618857627/Model_Card_Claude_3.pdf https://arxiv.org/pdf/2406.18521v1"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "claude-mythos-preview",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (The developer found the majority of CharXiv Reasoning question-answer text pairs in the full pretraining mix by exact-match search, even though evaluation images were filtered at image level. The evaluated items are 1,000 validation-split questions. On a 100-item manually perturbed remix, the developer judged memorisation effects limited.); this contradicts a statement that the data was excluded",
  "sources": "https://www-cdn.anthropic.com/08ab9158070959f88f296514c21b7facce6f52bc.pdf"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "gemini-1-5",
  "level": "not_seen",
  "reason": "its test items were published from 2024-06-26, after the model's stated training-data cutoff (2023-11-01)",
  "sources": "https://arxiv.org/pdf/2507.06261v6 https://arxiv.org/pdf/2406.18521v1"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "gemma-3n",
  "level": "not_seen",
  "reason": "its test items were published from 2024-06-26, after the model's stated training-data cutoff (2024-06-01)",
  "sources": "https://ai.google.dev/gemma/docs/gemma-3n/model_card https://arxiv.org/pdf/2406.18521v1"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "gpt-4",
  "level": "not_seen",
  "reason": "its test items were published from 2024-06-26, after the model's stated training-data cutoff (2021-09-01)",
  "sources": "https://arxiv.org/pdf/2303.08774v6 https://arxiv.org/pdf/2406.18521v1"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "gpt-4o",
  "level": "not_seen",
  "reason": "its test items were published from 2024-06-26, after the model's stated training-data cutoff (2023-10-01)",
  "sources": "https://arxiv.org/pdf/2410.21276v1 https://developers.openai.com/api/docs/models/gpt-4o https://arxiv.org/pdf/2406.18521v1"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "internvl3.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "internvl3",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "llama-3.2-vision",
  "level": "not_seen",
  "reason": "its test items were published from 2024-06-26, after the model's stated training-data cutoff (2023-12-01)",
  "sources": "https://raw.githubusercontent.com/meta-llama/llama-models/main/models/llama3_2/MODEL_CARD_VISION.md https://arxiv.org/pdf/2406.18521v1"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "llava-onevision-1.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "llava-onevision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "molmo",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "molmo2",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "o1",
  "level": "not_seen",
  "reason": "its test items were published from 2024-06-26, after the model's stated training-data cutoff (2023-10-01)",
  "sources": "https://developers.openai.com/api/docs/models/o1 https://arxiv.org/pdf/2406.18521v1"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "o3",
  "level": "not_seen",
  "reason": "its test items were published from 2024-06-26, after the model's stated training-data cutoff (2024-06-01)",
  "sources": "https://developers.openai.com/api/docs/models/o3 https://arxiv.org/pdf/2406.18521v1"
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-multimodal",
  "dataset": "charxiv",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "amber",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "bloom",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "claude-mythos-preview",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "dclm-7b-it",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "dclm-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "deepseek-r1-0528",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "deepseek-r1-distill",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "deepseek-r1",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2501.12948v2"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "deepseek-v3-0324",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "deepseek-v3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "deepseek-v3.2-exp",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "deepseek-v3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "deepseek-v3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "deepseek-v4.1-flash",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "deepseek-v4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "gemma-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "gpt-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "gpt-6-astra",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "gpt-oss",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "internvl3.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "internvl3",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "k2-65b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "k2-horizon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "k2-v2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "llama-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "llama-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "llama-3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "llama-3.3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "llama-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "llava-onevision-1.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "llava-onevision",
  "level": "unknown",
  "reason": "unknown for its base model SigLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "ministral-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "mistral-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "mixtral-8x7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "molmo",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "molmo2",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "olmo-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "olmo-2-1124",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "olmo-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "olmo-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "olmo-hybrid-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "phi-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "phi-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "phi-4-mini-flash-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "phi-4-mini-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "phi-4-mini",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "phi-4-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "phi-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "phi-slimmoe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "pythia-deduped",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "pythia",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "qwen2.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "qwen3-2507",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "qwen3-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "qwen3.6",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "qwen3.8-flash-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "qwen3.8",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "qwen3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "smollm2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "gpqa-diamond",
  "model": "smollm3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "amber",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "bloom",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "claude-mythos-preview",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "dclm-7b-it",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (16.20% of (downsampled) MMLU-Pro test questions are 'dirty' in DCLM-baseline.) [third-party study] (via DCLM-Baseline 1.0)",
  "sources": "https://huggingface.co/mlfoundations/dclm-7b-it/raw/main/README.md https://arxiv.org/pdf/2506.12229v4"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "dclm-7b",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (16.20% of (downsampled) MMLU-Pro test questions are 'dirty' in DCLM-baseline.) [third-party study] (via DCLM-Baseline 1.0)",
  "sources": "https://huggingface.co/apple/DCLM-7B/raw/main/README.md https://arxiv.org/pdf/2406.11794v4 https://arxiv.org/pdf/2506.12229v4"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "deepseek-r1-0528",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "deepseek-r1-distill",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "deepseek-r1",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2501.12948v2"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "deepseek-v3-0324",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "deepseek-v3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "deepseek-v3.2-exp",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "deepseek-v3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "deepseek-v3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "deepseek-v4.1-flash",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "deepseek-v4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "gemma-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "gpt-4",
  "level": "related",
  "reason": "a contamination analysis found test examples of the related dataset MMLU (Measuring Massive Multitask Language Understanding) (~0.6% of 1000 sampled MMLU test examples matched GPT-4 pre-training data (50-character substring match))",
  "sources": "https://arxiv.org/pdf/2303.08774v6 https://huggingface.co/datasets/TIGER-Lab/MMLU-Pro/raw/main/README.md https://arxiv.org/pdf/2406.01574v6"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "gpt-6-astra",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "gpt-oss",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "internvl3.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "internvl3",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "k2-65b",
  "level": "related",
  "reason": "a contamination analysis found test examples of the related dataset MMLU (Measuring Massive Multitask Language Understanding) (MMLU test questions are also substantially 'dirty' in Pile-train: DCLM-baseline (27.70%) has only 1.1x more dirty entries than Pile-train. Exact Pile-train rate is in Table 2 (not reliably extracted from the PDF text).) [third-party study] (via The Pile)",
  "sources": "https://huggingface.co/datasets/IFM/K2Datasets/raw/main/README.md https://arxiv.org/pdf/2506.12229v4 https://huggingface.co/datasets/TIGER-Lab/MMLU-Pro/raw/main/README.md https://arxiv.org/pdf/2406.01574v6"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "k2-horizon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "k2-v2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "llama-2",
  "level": "related",
  "reason": "a contamination analysis found test examples of the related dataset MMLU (Measuring Massive Multitask Language Understanding) (MMLU-Humanities (L=50): 709 of 4,705 samples (15%) 'not clean' (>=20% token contamination) and 520 (11%) 'dirty' (>=80%); flagged as boosted by contamination. MMLU-Overall for 70B: 2,180 of 14,042 (15.5%) not clean, 1,536 (10.9%) dirty, with a small effect on the overall score (-0.9 clean-vs-mean delta).)",
  "sources": "https://arxiv.org/pdf/2307.09288v2 https://arxiv.org/html/2307.09288v2 https://huggingface.co/datasets/TIGER-Lab/MMLU-Pro/raw/main/README.md https://arxiv.org/pdf/2406.01574v6"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "llama-3.1",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (8-gram overlap gave contamination scores too high to estimate a performance gain, even at higher thresholds; percentage not reported (dash in Table 15).)",
  "sources": "https://arxiv.org/pdf/2407.21783v3"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "llama-3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "llama-3.3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "llama-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "llava-onevision-1.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "llava-onevision",
  "level": "unknown",
  "reason": "unknown for its base model SigLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "ministral-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "mistral-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "mixtral-8x7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "molmo",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "molmo2",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "olmo-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "olmo-2-1124",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (16.20% of (downsampled) MMLU-Pro test questions are 'dirty' in DCLM-baseline.) [third-party study] (via DCLM-Baseline 1.0)",
  "sources": "https://huggingface.co/datasets/allenai/olmo-mix-1124/raw/main/README.md https://arxiv.org/pdf/2506.12229v4"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "olmo-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "olmo-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "olmo-hybrid-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "phi-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "phi-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "phi-4-mini-flash-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "phi-4-mini-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "phi-4-mini",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "phi-4-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "phi-4",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2412.08905v1"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "phi-slimmoe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "pythia-deduped",
  "level": "related",
  "reason": "a contamination analysis found test examples of the related dataset MMLU (Measuring Massive Multitask Language Understanding) (MMLU test questions are also substantially 'dirty' in Pile-train: DCLM-baseline (27.70%) has only 1.1x more dirty entries than Pile-train. Exact Pile-train rate is in Table 2 (not reliably extracted from the PDF text).) [third-party study] (via The Pile)",
  "sources": "https://huggingface.co/EleutherAI/pythia-12b-deduped/raw/main/README.md https://huggingface.co/EleutherAI/pythia-14m-deduped/raw/main/README.md https://huggingface.co/EleutherAI/pythia-31m-deduped/raw/main/README.md https://arxiv.org/pdf/2506.12229v4 https://huggingface.co/datasets/TIGER-Lab/MMLU-Pro/raw/main/README.md https://arxiv.org/pdf/2406.01574v6"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "pythia",
  "level": "related",
  "reason": "a contamination analysis found test examples of the related dataset MMLU (Measuring Massive Multitask Language Understanding) (MMLU test questions are also substantially 'dirty' in Pile-train: DCLM-baseline (27.70%) has only 1.1x more dirty entries than Pile-train. Exact Pile-train rate is in Table 2 (not reliably extracted from the PDF text).) [third-party study] (via The Pile)",
  "sources": "https://huggingface.co/EleutherAI/pythia-12b/raw/main/README.md https://huggingface.co/EleutherAI/pythia-14m/raw/main/README.md https://huggingface.co/EleutherAI/pythia-31m/raw/main/README.md https://arxiv.org/pdf/2506.12229v4 https://huggingface.co/datasets/TIGER-Lab/MMLU-Pro/raw/main/README.md https://arxiv.org/pdf/2406.01574v6"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "qwen2.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "qwen3-2507",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "qwen3-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "qwen3.6",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "qwen3.8-flash-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "qwen3.8",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "qwen3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "smollm2",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (16.20% of (downsampled) MMLU-Pro test questions are 'dirty' in DCLM-baseline.) [third-party study] (via DCLM-Baseline 1.0)",
  "sources": "https://arxiv.org/pdf/2502.02737v1 https://arxiv.org/pdf/2506.12229v4"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "mmlu-pro",
  "model": "smollm3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "amber",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "bloom",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "claude-mythos-preview",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "dclm-7b-it",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "dclm-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "deepseek-r1-0528",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "deepseek-r1-distill",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "deepseek-r1",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2501.12948v2"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "deepseek-v3-0324",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "deepseek-v3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "deepseek-v3.2-exp",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "deepseek-v3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "deepseek-v3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "deepseek-v4.1-flash",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "deepseek-v4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "gemma-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "gpt-4",
  "level": "not_seen",
  "reason": "its test items were published from 2023-05-01, after the model's stated training-data cutoff (2021-09-01)",
  "sources": "https://arxiv.org/pdf/2303.08774v6 https://raw.githubusercontent.com/LiveCodeBench/LiveCodeBench/main/README.md"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "gpt-6-astra",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "gpt-oss",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "internvl3.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "internvl3",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "k2-65b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "k2-horizon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "k2-v2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "llama-2",
  "level": "not_seen",
  "reason": "its test items were published from 2023-05-01, after the model's stated training-data cutoff (2022-09-01)",
  "sources": "https://raw.githubusercontent.com/meta-llama/llama/main/MODEL_CARD.md https://raw.githubusercontent.com/LiveCodeBench/LiveCodeBench/main/README.md"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "llama-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "llama-3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "llama-3.3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "llama-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "llava-onevision-1.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "llava-onevision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "ministral-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "mistral-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "mixtral-8x7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "molmo",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "molmo2",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "olmo-1",
  "level": "not_seen",
  "reason": "its test items were published from 2023-05-01, after the model's stated training-data cutoff (2023-02-01)",
  "sources": "https://huggingface.co/allenai/OLMo-7B/raw/main/README.md https://raw.githubusercontent.com/LiveCodeBench/LiveCodeBench/main/README.md"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "olmo-2-1124",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "olmo-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "olmo-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "olmo-hybrid-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "phi-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "phi-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "phi-4-mini-flash-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "phi-4-mini-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "phi-4-mini",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "phi-4-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "phi-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "phi-slimmoe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "pythia-deduped",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "pythia",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "qwen2.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "qwen3-2507",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "qwen3-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "qwen3.6",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "qwen3.8-flash-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "qwen3.8",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "qwen3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "smollm2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "livecodebench",
  "model": "smollm3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "amber",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "bloom",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "claude-mythos-preview",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the training data (Developer memorisation audit found some signs of memorisation on SWE-bench Verified, Multilingual and Pro despite corpus-level decontamination; a high-recall filter (threshold 0.7) removes 8-15% of each benchmark)",
  "sources": "https://www-cdn.anthropic.com/08ab9158070959f88f296514c21b7facce6f52bc.pdf"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "claude-opus-4-5",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the training data (Third-party (OpenAI) elicitation study found strong contamination examples: Claude Opus 4.5 reproduced SWE-bench Verified gold-patch / task-specific details) [third-party study]",
  "sources": "https://openai.com/index/why-we-no-longer-evaluate-swe-bench-verified/"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "dclm-7b-it",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "dclm-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "deepseek-r1-0528",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "deepseek-r1-distill",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "deepseek-r1",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2501.12948v2"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "deepseek-v3-0324",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "deepseek-v3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "deepseek-v3.2-exp",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "deepseek-v3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "deepseek-v3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "deepseek-v4.1-flash",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "deepseek-v4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "gemma-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "gpt-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "gpt-5-2",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the training data (Developer analysis found GPT-5.2 (and GPT-5.2-Chat) could reproduce SWE-bench Verified gold patches / task details and solved 31 tasks judged almost impossible without prior exposure; OpenAI stopped reporting SWE-bench Verified)",
  "sources": "https://openai.com/index/why-we-no-longer-evaluate-swe-bench-verified/"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "gpt-6-astra",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "gpt-oss",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "internvl3.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "internvl3",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "k2-65b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "k2-horizon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "k2-v2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "llama-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "llama-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "llama-3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "llama-3.3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "llama-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "llava-onevision-1.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "llava-onevision",
  "level": "unknown",
  "reason": "unknown for its base model SigLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "ministral-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "mistral-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "mixtral-8x7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "molmo",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "molmo2",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "o3",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the training data (Identifies the buggy file path from the issue text alone in up to 76% of SWE-bench Verified tasks, vs <53% on tasks from other repositories; the authors read it as possible memorisation/contamination.) [third-party study]",
  "sources": "https://arxiv.org/pdf/2506.12286v4"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "olmo-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "olmo-2-1124",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "olmo-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "olmo-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "olmo-hybrid-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "phi-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "phi-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "phi-4-mini-flash-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "phi-4-mini-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "phi-4-mini",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "phi-4-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "phi-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "phi-slimmoe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "pythia-deduped",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "pythia",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "qwen2.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "qwen3-2507",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "qwen3-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "qwen3.6",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "qwen3.8-flash-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "qwen3.8",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "qwen3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "smollm2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "swe-bench-verified",
  "model": "smollm3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "amber",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "bloom",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "claude-3",
  "level": "not_seen",
  "reason": "its test items were published from 2024-01-31, after the model's stated training-data cutoff (2023-08-01)",
  "sources": "https://www-cdn.anthropic.com/de8ba9b01c9ab7cbabf5c33b80b7bbc618857627/Model_Card_Claude_3.pdf https://web.archive.org/web/20240302025113/https://artofproblemsolving.com/wiki/index.php/2024_AIME_I https://web.archive.org/web/20240227182215/https://maa.org/node/79/"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "claude-mythos-preview",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "dclm-7b-it",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "dclm-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "deepseek-r1-0528",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "deepseek-r1-distill",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "deepseek-r1",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2501.12948v2"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "deepseek-v3-0324",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "deepseek-v3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "deepseek-v3.2-exp",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "deepseek-v3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "deepseek-v3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "deepseek-v4.1-flash",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "deepseek-v4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "gemini-1-5",
  "level": "not_seen",
  "reason": "its test items were published from 2024-01-31, after the model's stated training-data cutoff (2023-11-01)",
  "sources": "https://arxiv.org/pdf/2507.06261v6 https://web.archive.org/web/20240302025113/https://artofproblemsolving.com/wiki/index.php/2024_AIME_I https://web.archive.org/web/20240227182215/https://maa.org/node/79/"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "gemma-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "gpt-4",
  "level": "not_seen",
  "reason": "its test items were published from 2024-01-31, after the model's stated training-data cutoff (2021-09-01)",
  "sources": "https://arxiv.org/pdf/2303.08774v6 https://web.archive.org/web/20240302025113/https://artofproblemsolving.com/wiki/index.php/2024_AIME_I https://web.archive.org/web/20240227182215/https://maa.org/node/79/"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "gpt-4o",
  "level": "not_seen",
  "reason": "its test items were published from 2024-01-31, after the model's stated training-data cutoff (2023-10-01)",
  "sources": "https://arxiv.org/pdf/2410.21276v1 https://developers.openai.com/api/docs/models/gpt-4o https://web.archive.org/web/20240302025113/https://artofproblemsolving.com/wiki/index.php/2024_AIME_I https://web.archive.org/web/20240227182215/https://maa.org/node/79/"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "gpt-6-astra",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "gpt-oss",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "internvl3.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "internvl3",
  "level": "test_seen",
  "reason": "via base model Qwen2.5: a contamination analysis found test examples in the pre-training data (Exact completion of 13.33% of AIME 2024 problems from a 60% prefix (Table 2).) [third-party study]",
  "sources": "https://arxiv.org/pdf/2504.10479v3 https://arxiv.org/pdf/2507.10532v3"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "k2-65b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "k2-horizon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "k2-v2",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (Sentence-level memorisation rate 23.32% for K2 base (beam-search continuation exactly reproduces the next sentence of the question); lower than Qwen2.5-72B (41.96%) and Olmo 3 32B (37.82%); authors attribute AIME performance to reasoning rather than memorisation.)",
  "sources": "https://arxiv.org/pdf/2512.06201v3"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "llama-2",
  "level": "not_seen",
  "reason": "its test items were published from 2024-01-31, after the model's stated training-data cutoff (2022-09-01)",
  "sources": "https://raw.githubusercontent.com/meta-llama/llama/main/MODEL_CARD.md https://web.archive.org/web/20240302025113/https://artofproblemsolving.com/wiki/index.php/2024_AIME_I https://web.archive.org/web/20240227182215/https://maa.org/node/79/"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "llama-3.1",
  "level": "not_seen",
  "reason": "its test items were published from 2024-01-31, after the model's stated training-data cutoff (2023-12-01)",
  "sources": "https://raw.githubusercontent.com/meta-llama/llama-models/main/models/llama3_1/MODEL_CARD.md https://arxiv.org/pdf/2407.21783v3 https://web.archive.org/web/20240302025113/https://artofproblemsolving.com/wiki/index.php/2024_AIME_I https://web.archive.org/web/20240227182215/https://maa.org/node/79/"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "llama-3.2-vision",
  "level": "not_seen",
  "reason": "its test items were published from 2024-01-31, after the model's stated training-data cutoff (2023-12-01)",
  "sources": "https://raw.githubusercontent.com/meta-llama/llama-models/main/models/llama3_2/MODEL_CARD_VISION.md https://web.archive.org/web/20240302025113/https://artofproblemsolving.com/wiki/index.php/2024_AIME_I https://web.archive.org/web/20240227182215/https://maa.org/node/79/"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "llama-3.2",
  "level": "not_seen",
  "reason": "its test items were published from 2024-01-31, after the model's stated training-data cutoff (2023-12-01)",
  "sources": "https://raw.githubusercontent.com/meta-llama/llama-models/main/models/llama3_2/MODEL_CARD.md https://web.archive.org/web/20240302025113/https://artofproblemsolving.com/wiki/index.php/2024_AIME_I https://web.archive.org/web/20240227182215/https://maa.org/node/79/"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "llama-3.3",
  "level": "not_seen",
  "reason": "its test items were published from 2024-01-31, after the model's stated training-data cutoff (2023-12-01)",
  "sources": "https://raw.githubusercontent.com/meta-llama/llama-models/main/models/llama3_3/MODEL_CARD.md https://web.archive.org/web/20240302025113/https://artofproblemsolving.com/wiki/index.php/2024_AIME_I https://web.archive.org/web/20240227182215/https://maa.org/node/79/"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "llama-3",
  "level": "not_seen",
  "reason": "its test items were published from 2024-01-31, after the model's stated training-data cutoff (2023-12-01)",
  "sources": "https://raw.githubusercontent.com/meta-llama/llama3/main/MODEL_CARD.md https://web.archive.org/web/20240302025113/https://artofproblemsolving.com/wiki/index.php/2024_AIME_I https://web.archive.org/web/20240227182215/https://maa.org/node/79/"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "llava-onevision-1.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "llava-onevision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "ministral-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "mistral-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "mixtral-8x7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "molmo",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "molmo2",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "o1",
  "level": "not_seen",
  "reason": "its test items were published from 2024-01-31, after the model's stated training-data cutoff (2023-10-01)",
  "sources": "https://developers.openai.com/api/docs/models/o1 https://web.archive.org/web/20240302025113/https://artofproblemsolving.com/wiki/index.php/2024_AIME_I https://web.archive.org/web/20240227182215/https://maa.org/node/79/"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "olmo-1",
  "level": "not_seen",
  "reason": "its test items were published from 2024-01-31, after the model's stated training-data cutoff (2023-02-01)",
  "sources": "https://huggingface.co/allenai/OLMo-7B/raw/main/README.md https://web.archive.org/web/20240302025113/https://artofproblemsolving.com/wiki/index.php/2024_AIME_I https://web.archive.org/web/20240227182215/https://maa.org/node/79/"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "olmo-2-1124",
  "level": "not_seen",
  "reason": "its test items were published from 2024-01-31, after the model's stated training-data cutoff (2023-12-01)",
  "sources": "https://huggingface.co/allenai/OLMo-2-1124-7B/raw/main/README.md https://web.archive.org/web/20240302025113/https://artofproblemsolving.com/wiki/index.php/2024_AIME_I https://web.archive.org/web/20240227182215/https://maa.org/node/79/"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "olmo-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "olmo-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "olmo-hybrid-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "phi-3.5",
  "level": "not_seen",
  "reason": "its test items were published from 2024-01-31, after the model's stated training-data cutoff (2023-10-01)",
  "sources": "https://huggingface.co/microsoft/Phi-3.5-mini-instruct/raw/main/README.md https://web.archive.org/web/20240302025113/https://artofproblemsolving.com/wiki/index.php/2024_AIME_I https://web.archive.org/web/20240227182215/https://maa.org/node/79/"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "phi-3",
  "level": "not_seen",
  "reason": "its test items were published from 2024-01-31, after the model's stated training-data cutoff (2023-10-01)",
  "sources": "https://huggingface.co/microsoft/Phi-3-medium-4k-instruct/raw/main/README.md https://web.archive.org/web/20240302025113/https://artofproblemsolving.com/wiki/index.php/2024_AIME_I https://web.archive.org/web/20240227182215/https://maa.org/node/79/"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "phi-4-mini-flash-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "phi-4-mini-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "phi-4-mini",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "phi-4-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "phi-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "phi-slimmoe",
  "level": "not_seen",
  "reason": "its test items were published from 2024-01-31, after the model's stated training-data cutoff (2023-10-01)",
  "sources": "https://huggingface.co/microsoft/Phi-mini-MoE-instruct/raw/main/README.md https://web.archive.org/web/20240302025113/https://artofproblemsolving.com/wiki/index.php/2024_AIME_I https://web.archive.org/web/20240227182215/https://maa.org/node/79/"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "pythia-deduped",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "pythia",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "qwen2.5-vl",
  "level": "test_seen",
  "reason": "via base model Qwen2.5: a contamination analysis found test examples in the pre-training data (Exact completion of 13.33% of AIME 2024 problems from a 60% prefix (Table 2).) [third-party study]",
  "sources": "https://arxiv.org/pdf/2502.13923v1 https://arxiv.org/pdf/2507.10532v3"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "qwen2.5",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (Exact completion of 13.33% of AIME 2024 problems from a 60% prefix (Table 2).) [third-party study]",
  "sources": "https://arxiv.org/pdf/2507.10532v3"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "qwen3-2507",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "qwen3-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "qwen3.6",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "qwen3.8-flash-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "qwen3.8",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "qwen3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "smollm2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2024",
  "model": "smollm3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "amber",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "bloom",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "claude-3-5-sonnet",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2024-04-01)",
  "sources": "https://web.archive.org/web/20241031205657id_/https://docs.anthropic.com/en/docs/about-claude/models https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "claude-3-7-sonnet",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2024-11-01)",
  "sources": "https://assets.anthropic.com/m/785e231869ea8b3b/original/claude-3-7-sonnet-system-card.pdf https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "claude-3",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2023-08-01)",
  "sources": "https://www-cdn.anthropic.com/de8ba9b01c9ab7cbabf5c33b80b7bbc618857627/Model_Card_Claude_3.pdf https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "claude-mythos-preview",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "claude-opus-4-5",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the training data (Developer found rephrased AIME questions, official solutions and model answers in the training corpus despite decontamination; the AIME 2025 score may be inflated (example answer suggesting memorisation))",
  "sources": "https://assets.anthropic.com/m/64823ba7485345a7/Claude-Opus-4-5-System-Card.pdf"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "dclm-7b-it",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "dclm-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "deepseek-r1-0528",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "deepseek-r1-distill",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "deepseek-r1",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2501.12948v2"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "deepseek-v3-0324",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "deepseek-v3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "deepseek-v3.2-exp",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "deepseek-v3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "deepseek-v3",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2024-07-01)",
  "sources": "https://arxiv.org/pdf/2501.12948v2 https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "deepseek-v4.1-flash",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "deepseek-v4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "gemini-1-5",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2023-11-01)",
  "sources": "https://arxiv.org/pdf/2507.06261v6 https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "gemini-2-5",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2025-01-01)",
  "sources": "https://arxiv.org/pdf/2507.06261v6 https://ai.google.dev/gemini-api/docs/models/gemini-2.5-pro https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "gemini-3-pro",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2025-01-01)",
  "sources": "https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-3-Pro-Model-Card.pdf https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "gemma-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "gemma-3",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2024-08-01)",
  "sources": "https://ai.google.dev/gemma/docs/core/model_card_3 https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "gemma-3n",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2024-06-01)",
  "sources": "https://ai.google.dev/gemma/docs/gemma-3n/model_card https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "gemma-4",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2025-01-01)",
  "sources": "https://arxiv.org/pdf/2607.02770v2 https://ai.google.dev/gemma/docs/core/model_card_4 https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "gpt-4",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2021-09-01)",
  "sources": "https://arxiv.org/pdf/2303.08774v6 https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "gpt-4o",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2023-10-01)",
  "sources": "https://arxiv.org/pdf/2410.21276v1 https://developers.openai.com/api/docs/models/gpt-4o https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "gpt-5",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2024-09-30)",
  "sources": "https://developers.openai.com/api/docs/models/gpt-5 https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "gpt-6-astra",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "gpt-oss",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2024-06-01)",
  "sources": "https://arxiv.org/pdf/2508.10925v1 https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "internvl3.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "internvl3",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "k2-65b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "k2-horizon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "k2-v2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "llama-2",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2022-09-01)",
  "sources": "https://raw.githubusercontent.com/meta-llama/llama/main/MODEL_CARD.md https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "llama-3.1",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2023-12-01)",
  "sources": "https://raw.githubusercontent.com/meta-llama/llama-models/main/models/llama3_1/MODEL_CARD.md https://arxiv.org/pdf/2407.21783v3 https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "llama-3.2-vision",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2023-12-01)",
  "sources": "https://raw.githubusercontent.com/meta-llama/llama-models/main/models/llama3_2/MODEL_CARD_VISION.md https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "llama-3.2",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2023-12-01)",
  "sources": "https://raw.githubusercontent.com/meta-llama/llama-models/main/models/llama3_2/MODEL_CARD.md https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "llama-3.3",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2023-12-01)",
  "sources": "https://raw.githubusercontent.com/meta-llama/llama-models/main/models/llama3_3/MODEL_CARD.md https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "llama-3",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2023-12-01)",
  "sources": "https://raw.githubusercontent.com/meta-llama/llama3/main/MODEL_CARD.md https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "llama-4",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2024-08-01)",
  "sources": "https://raw.githubusercontent.com/meta-llama/llama-models/main/models/llama4/MODEL_CARD.md https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "llava-onevision-1.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "llava-onevision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "ministral-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "mistral-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "mistral-medium-3.5",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2024-11-01)",
  "sources": "https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/raw/main/SYSTEM_PROMPT.txt https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "mistral-small-4",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2024-11-01)",
  "sources": "https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/raw/main/SYSTEM_PROMPT.txt https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "mixtral-8x7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "molmo",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "molmo2",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "o1",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2023-10-01)",
  "sources": "https://developers.openai.com/api/docs/models/o1 https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "o3",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2024-06-01)",
  "sources": "https://developers.openai.com/api/docs/models/o3 https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "olmo-1",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2023-02-01)",
  "sources": "https://huggingface.co/allenai/OLMo-7B/raw/main/README.md https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "olmo-2-1124",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2023-12-01)",
  "sources": "https://huggingface.co/allenai/OLMo-2-1124-7B/raw/main/README.md https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "olmo-3.1",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2024-12-01)",
  "sources": "https://huggingface.co/allenai/Olmo-3.1-32B-Think/raw/main/README.md https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "olmo-3",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2024-12-01)",
  "sources": "https://huggingface.co/allenai/Olmo-3-1025-7B/raw/main/README.md https://arxiv.org/pdf/2512.13961v2 https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "olmo-hybrid-7b",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2024-12-01)",
  "sources": "https://huggingface.co/allenai/Olmo-Hybrid-7B/raw/main/README.md https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "phi-3.5",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2023-10-01)",
  "sources": "https://huggingface.co/microsoft/Phi-3.5-mini-instruct/raw/main/README.md https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "phi-3",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2023-10-01)",
  "sources": "https://huggingface.co/microsoft/Phi-3-medium-4k-instruct/raw/main/README.md https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "phi-4-mini-flash-reasoning",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2025-02-01)",
  "sources": "https://huggingface.co/microsoft/Phi-4-mini-flash-reasoning/raw/main/README.md https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "phi-4-mini-reasoning",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2025-02-01)",
  "sources": "https://huggingface.co/microsoft/Phi-4-mini-reasoning/raw/main/README.md https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "phi-4-mini",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2024-06-01)",
  "sources": "https://huggingface.co/microsoft/Phi-4-mini-instruct/raw/main/README.md https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "phi-4-reasoning",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2504.21318v1"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "phi-4",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2024-06-01)",
  "sources": "https://huggingface.co/microsoft/phi-4/raw/main/README.md https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "phi-slimmoe",
  "level": "not_seen",
  "reason": "its test items were published from 2025-02-06, after the model's stated training-data cutoff (2023-10-01)",
  "sources": "https://huggingface.co/microsoft/Phi-mini-MoE-instruct/raw/main/README.md https://web.archive.org/web/20250501204857/https://artofproblemsolving.com/wiki/index.php/2025_AIME_I https://maa.org/?p=8152"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "pythia-deduped",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "pythia",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "qwen2.5-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen2.5: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "qwen2.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "qwen3-2507",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "qwen3-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "qwen3.6",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "qwen3.8-flash-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "qwen3.8",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "qwen3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "smollm2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "aime-2025",
  "model": "smollm3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "amber",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "bloom",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "claude-3-5-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "claude-3-7-sonnet",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "claude-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "claude-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "claude-fable-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "claude-mythos-preview",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "claude-opus-4-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "claude-opus-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "claude-sonnet-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "dclm-7b-it",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "dclm-7b",
  "level": "related",
  "reason": "a contamination analysis found test examples of the related dataset MATH (Mathematics Aptitude Test of Heuristics) (Via ProofPile2 (Llemma paper): ~7% of MATH test problem statements and 0.6% of MATH test solutions have 30-gram hits in Proof-Pile-2 (OpenWebMath/AlgebraicStack). CONDA 2024 report also lists hendrycks/competition_math and gsm8k as contaminated in EleutherAI/proof-pile-2.) [third-party study]",
  "sources": "https://arxiv.org/pdf/2310.10631v3 https://arxiv.org/pdf/2407.21530v2 https://raw.githubusercontent.com/openai/prm800k/main/README.md https://huggingface.co/datasets/HuggingFaceH4/MATH-500/raw/main/README.md"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "deepseek-r1-0528",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "deepseek-r1-distill",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "deepseek-r1",
  "level": "not_seen",
  "reason": "the authors state its test portion was excluded from training",
  "sources": "https://arxiv.org/pdf/2501.12948v2"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "deepseek-v3-0324",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "deepseek-v3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "deepseek-v3.2-exp",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "deepseek-v3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "deepseek-v3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "deepseek-v4.1-flash",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "deepseek-v4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "gemini-1-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "gemini-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "gemini-3-pro",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "gemini-4-argon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "gemma-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "gemma-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "gemma-3n",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "gemma-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "gpt-4",
  "level": "related",
  "reason": "related dataset MATH (Mathematics Aptitude Test of Heuristics) was used",
  "sources": "https://arxiv.org/pdf/2303.08774v6 https://raw.githubusercontent.com/openai/prm800k/main/README.md https://huggingface.co/datasets/HuggingFaceH4/MATH-500/raw/main/README.md"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "gpt-4o",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "gpt-5-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "gpt-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "gpt-6-astra",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "gpt-oss",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "internvl3.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "internvl3",
  "level": "test_seen",
  "reason": "via base model Qwen2.5: a contamination analysis found test examples in the pre-training data (Exact completion of 21.2% of MATH-500 problems from a 60% prefix (Table 2).) [third-party study]",
  "sources": "https://arxiv.org/pdf/2504.10479v3 https://arxiv.org/pdf/2507.10532v3"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "k2-65b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "k2-horizon",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "k2-v2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "llama-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "llama-3.1",
  "level": "related",
  "reason": "a contamination analysis found test examples of the related dataset MATH (Mathematics Aptitude Test of Heuristics) (1% of MATH examples considered contaminated at the selected threshold; estimated gain 0.0 / -0.1 / -0.2. Authors note low thresholds yield high contamination for MATH but no performance gains.)",
  "sources": "https://arxiv.org/html/2407.21783v3 https://raw.githubusercontent.com/openai/prm800k/main/README.md https://huggingface.co/datasets/HuggingFaceH4/MATH-500/raw/main/README.md"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "llama-3.2-vision",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "llama-3.2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "llama-3.3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "llama-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "llama-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "llava-onevision-1.5",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "llava-onevision",
  "level": "unknown",
  "reason": "unknown for its base model SigLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "ministral-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "mistral-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "mistral-large-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "mistral-medium-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "mistral-small-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "mixtral-8x7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "molmo",
  "level": "unknown",
  "reason": "unknown for its base model OpenAI CLIP: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "molmo2",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "o1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "o3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "olmo-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "olmo-2-1124",
  "level": "related",
  "reason": "in fine-tuning data: related dataset MATH (Mathematics Aptitude Test of Heuristics) was used",
  "sources": "https://arxiv.org/pdf/2501.00656v3 https://raw.githubusercontent.com/openai/prm800k/main/README.md https://huggingface.co/datasets/HuggingFaceH4/MATH-500/raw/main/README.md"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "olmo-3.1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "olmo-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "olmo-hybrid-7b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "phi-3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "phi-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "phi-4-mini-flash-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "phi-4-mini-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "phi-4-mini",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "phi-4-reasoning",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "phi-4",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "phi-slimmoe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "pixtral-12b",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "pixtral-large",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "pythia-deduped",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "pythia",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "qwen2.5-vl",
  "level": "test_seen",
  "reason": "via base model Qwen2.5: a contamination analysis found test examples in the pre-training data (Exact completion of 21.2% of MATH-500 problems from a 60% prefix (Table 2).) [third-party study]",
  "sources": "https://arxiv.org/pdf/2502.13923v1 https://arxiv.org/pdf/2507.10532v3"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "qwen2.5",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (Exact completion of 21.2% of MATH-500 problems from a 60% prefix (Table 2).) [third-party study]",
  "sources": "https://arxiv.org/pdf/2507.10532v3"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "qwen3-2507",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "qwen3-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "qwen3-vl",
  "level": "unknown",
  "reason": "unknown for its base model Qwen3: training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "qwen3.5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "qwen3.6",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "qwen3.8-flash-next",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "qwen3.8",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "qwen3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "smollm2",
  "level": "related",
  "reason": "in fine-tuning data: related dataset MATH (Mathematics Aptitude Test of Heuristics) was used (via SmolTalk)",
  "sources": "https://arxiv.org/pdf/2502.02737v1 https://huggingface.co/HuggingFaceTB/SmolLM2-1.7B-Instruct/raw/main/README.md https://huggingface.co/datasets/HuggingFaceTB/smoltalk/raw/main/README.md https://huggingface.co/datasets/meta-math/MetaMathQA/raw/main/README.md https://raw.githubusercontent.com/openai/prm800k/main/README.md https://huggingface.co/datasets/HuggingFaceH4/MATH-500/raw/main/README.md"
 },
 {
  "benchmark": "llm-recent",
  "dataset": "math-500",
  "model": "smollm3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "lag-llama",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "moirai-2.0-r",
  "level": "related",
  "reason": "related dataset ETT1 (GIFT-Eval) was used",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2205.13504v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "moment-1",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data (via Time Series Pile); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "time-moe",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "timer-s1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "timer-xl",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2410.04803v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "timer",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "timesfm-1-0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/gifteval/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "toto-1.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "toto-2.0-2.5b-ft",
  "level": "related",
  "reason": "related dataset ETT1 (GIFT-Eval) was used",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2 https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2205.13504v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "ttm-r3",
  "level": "related",
  "reason": "related dataset ETT1 (GIFT-Eval) was used",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2205.13504v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "units-few-shot-anomaly-detection",
  "level": "train_seen",
  "reason": "via base model UniTS: only the training split was used, not the benchmark's test data",
  "sources": "https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth1",
  "model": "units",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "lag-llama",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "moirai-2.0-r",
  "level": "related",
  "reason": "related dataset ETT2 (GIFT-Eval) was used",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2205.13504v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "moment-1",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data (via Time Series Pile); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "time-moe",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "timer-s1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "timer-xl",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2410.04803v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "timer",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "timesfm-1-0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/gifteval/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "toto-1.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "toto-2.0-2.5b-ft",
  "level": "related",
  "reason": "related dataset ETT2 (GIFT-Eval) was used",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2 https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2205.13504v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "ttm-r3",
  "level": "related",
  "reason": "related dataset ETT2 (GIFT-Eval) was used",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2205.13504v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "etth2",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "lag-llama",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "moirai-2.0-r",
  "level": "related",
  "reason": "related dataset ETT1 (GIFT-Eval) was used",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2205.13504v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "moment-1",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data (via Time Series Pile); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "time-moe",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "timer-s1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "timer-xl",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2410.04803v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "timer",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "timesfm-1-0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/gifteval/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "toto-1.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "toto-2.0-2.5b-ft",
  "level": "related",
  "reason": "related dataset ETT1 (GIFT-Eval) was used",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2 https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2205.13504v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "ttm-r3",
  "level": "related",
  "reason": "related dataset ETT1 (GIFT-Eval) was used",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2205.13504v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "units-few-shot-anomaly-detection",
  "level": "related",
  "reason": "via base model UniTS: related dataset ETTh1 was used",
  "sources": "https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py https://raw.githubusercontent.com/zhouhaoyi/ETDataset/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm1",
  "model": "units",
  "level": "related",
  "reason": "related dataset ETTh1 was used",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py https://raw.githubusercontent.com/zhouhaoyi/ETDataset/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "lag-llama",
  "level": "related",
  "reason": "related dataset ETTh2 was used; the authors state this dataset was excluded",
  "sources": "https://arxiv.org/pdf/2310.08278v3 https://raw.githubusercontent.com/zhouhaoyi/ETDataset/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "moirai-2.0-r",
  "level": "related",
  "reason": "related dataset ETT2 (GIFT-Eval) was used",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2205.13504v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "moment-1",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data (via Time Series Pile); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "time-moe",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "timer-s1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "timer-xl",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2410.04803v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "timer",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "timesfm-1-0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/gifteval/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "toto-1.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "toto-2.0-2.5b-ft",
  "level": "related",
  "reason": "related dataset ETT2 (GIFT-Eval) was used",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2 https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2205.13504v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "ttm-r3",
  "level": "related",
  "reason": "related dataset ETT2 (GIFT-Eval) was used",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2205.13504v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ettm2",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "chronos-2",
  "level": "related",
  "reason": "related dataset ElectricityLoadDiagrams20112014 (UCI) was used",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://raw.githubusercontent.com/laiguokun/multivariate-time-series-data/master/README.md https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "chronos-bolt",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))",
  "sources": "https://github.com/amazon-science/chronos-forecasting/issues/245 https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "chronos-t5",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via Chronos training corpus (real-world datasets))",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "lag-llama",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "moirai-2.0-r",
  "level": "related",
  "reason": "related dataset ElectricityLoadDiagrams20112014 (UCI) was used",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md https://raw.githubusercontent.com/laiguokun/multivariate-time-series-data/master/README.md https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "moment-1",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data (via Time Series Pile); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "time-moe",
  "level": "related",
  "reason": "related dataset ElectricityLoadDiagrams20112014 (UCI) was used (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4 https://raw.githubusercontent.com/laiguokun/multivariate-time-series-data/master/README.md https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "timer-s1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "timer-xl",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2410.04803v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "timer",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "timesfm-1-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/gifteval/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "tirex",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "toto-1.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "toto-2.0-2.5b-ft",
  "level": "related",
  "reason": "related dataset ElectricityLoadDiagrams20112014 (UCI) was used",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2 https://raw.githubusercontent.com/laiguokun/multivariate-time-series-data/master/README.md https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "ttm-r3",
  "level": "related",
  "reason": "related dataset ElectricityLoadDiagrams20112014 (UCI) was used",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://raw.githubusercontent.com/laiguokun/multivariate-time-series-data/master/README.md https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "units-few-shot-anomaly-detection",
  "level": "train_seen",
  "reason": "via base model UniTS: only the training split was used, not the benchmark's test data",
  "sources": "https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "electricity-321",
  "model": "units",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "chronos-2",
  "level": "related",
  "reason": "related dataset LargeST (PeMS California) was used",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "granite-ttm-r3",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via GIFT-Eval Pretrain (GiftEvalPretrain))",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "lag-llama",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "moirai-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "moirai-2.0-r",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via GIFT-Eval Pretrain (GiftEvalPretrain))",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "moirai-moe-1.0-r",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "moment-1",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data (via Time Series Pile); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "sundial",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "timer-s1",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive)); this contradicts a statement that the data was excluded",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "timesfm-1-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "timesfm-2-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2310.10688v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "timesfm-2-5",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via GIFT-Eval Pretrain (GiftEvalPretrain))",
  "sources": "https://huggingface.co/google/timesfm-2.5-200m-pytorch/raw/main/README.md https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "timesfm-3-0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via GIFT-Eval Pretrain (GiftEvalPretrain))",
  "sources": "https://huggingface.co/google/timesfm-3.0-pytorch/raw/main/README.md https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "tirex-1.1-gifteval",
  "level": "related",
  "reason": "related dataset LargeST (PeMS California) was used",
  "sources": "https://arxiv.org/pdf/2505.23719v2 https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "tirex-2-gifteval-pretrain",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via GIFT-Eval Pretrain (GiftEvalPretrain))",
  "sources": "https://huggingface.co/NX-AI/TiRex-2/raw/main/README.md https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "tirex-2-gifteval-zs",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "tirex",
  "level": "related",
  "reason": "related dataset LargeST (PeMS California) was used",
  "sources": "https://arxiv.org/pdf/2505.23719v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "toto-1.0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via GIFT-Eval Pretrain (GiftEvalPretrain))",
  "sources": "https://arxiv.org/pdf/2505.14766v2 https://huggingface.co/Datadog/Toto-Open-Base-1.0/raw/main/README.md https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "toto-2.0-2.5b-ft",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via GIFT-Eval Pretrain (GiftEvalPretrain))",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "toto-2024",
  "level": "seen_split_unclear",
  "reason": "a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "ttm-r3",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via GIFT-Eval Pretrain (GiftEvalPretrain))",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "units-few-shot-anomaly-detection",
  "level": "train_seen",
  "reason": "via base model UniTS: only the training split was used, not the benchmark's test data",
  "sources": "https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "traffic-862",
  "model": "units",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2402.02592v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "moirai-2.0-r",
  "level": "related",
  "reason": "related dataset Jena Weather (GIFT-Eval) was used",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "moment-1",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data (via Time Series Pile); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "sundial",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2502.00816v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "time-moe",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "timer-s1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "timer-xl",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2410.04803v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "timer",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/gifteval/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "toto-1.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "toto-2.0-2.5b-ft",
  "level": "related",
  "reason": "related dataset Jena Weather (GIFT-Eval) was used",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2605.20119v2 https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "ttm-r3",
  "level": "related",
  "reason": "related dataset Jena Weather (GIFT-Eval) was used",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "units-few-shot-anomaly-detection",
  "level": "train_seen",
  "reason": "via base model UniTS: only the training split was used, not the benchmark's test data",
  "sources": "https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "weather-jena-2020",
  "model": "units",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2403.07815v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "moirai-2.0-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "moment-1",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data (via Time Series Pile); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "sundial",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "time-moe",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "timer-s1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "tirex-2-fevbench",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "tirex-2-gifteval-zs",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2607.01204v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "toto-1.0",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2505.14766v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "toto-2.0-2.5b-ft",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "units-few-shot-anomaly-detection",
  "level": "train_seen",
  "reason": "via base model UniTS: only the training split was used, not the benchmark's test data",
  "sources": "https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "exchange-rate",
  "model": "units",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "granite-ttm-r2-1",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md https://arxiv.org/pdf/2106.13008v5 https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "granite-ttm-r3",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used (via GIFT-Eval Pretrain (GiftEvalPretrain))",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r3/raw/main/README.md https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "moirai-1.0-r",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://raw.githubusercontent.com/SalesforceAIResearch/uni2ts/main/project/moirai-1/README.md https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md https://arxiv.org/pdf/2106.13008v5 https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "moirai-2.0-r",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used (via GIFT-Eval Pretrain (GiftEvalPretrain))",
  "sources": "https://arxiv.org/pdf/2511.11698v3 https://huggingface.co/Salesforce/moirai-2.0-R-small/raw/main/README.md https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "moirai-moe-1.0-r",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.10469v1 https://arxiv.org/pdf/2402.02592v2 https://arxiv.org/pdf/2106.13008v5 https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "moment-1",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data (via Time Series Pile); the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "sundial",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "time-moe",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used (via Time-300B)",
  "sources": "https://arxiv.org/pdf/2409.16040v4 https://arxiv.org/pdf/2106.13008v5 https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "timer-s1",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2603.04791v3 https://arxiv.org/pdf/2402.02592v2 https://arxiv.org/pdf/2106.13008v5 https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "timer-xl",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://arxiv.org/pdf/2402.02592v2 https://arxiv.org/pdf/2106.13008v5 https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "timer",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://arxiv.org/pdf/2402.02592v2 https://arxiv.org/pdf/2106.13008v5 https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "timesfm-2-0",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://huggingface.co/google/timesfm-2.0-500m-pytorch/raw/main/README.md https://arxiv.org/pdf/2402.02592v2 https://arxiv.org/pdf/2106.13008v5 https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "timesfm-2-5",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used (via GIFT-Eval Pretrain (GiftEvalPretrain))",
  "sources": "https://huggingface.co/google/timesfm-2.5-200m-pytorch/raw/main/README.md https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "timesfm-3-0",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used (via GIFT-Eval Pretrain (GiftEvalPretrain))",
  "sources": "https://huggingface.co/google/timesfm-3.0-pytorch/raw/main/README.md https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "tirex-1.1-gifteval",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used",
  "sources": "https://arxiv.org/pdf/2505.23719v2 https://huggingface.co/NX-AI/TiRex-1.1-gifteval/raw/main/README.md https://arxiv.org/pdf/2106.13008v5 https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "tirex-2-fevbench",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://arxiv.org/pdf/2106.13008v5 https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "tirex-2-gifteval-pretrain",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used (via GIFT-Eval Pretrain (GiftEvalPretrain))",
  "sources": "https://huggingface.co/NX-AI/TiRex-2/raw/main/README.md https://huggingface.co/NX-AI/TiRex-2-gifteval-pretrain/raw/main/README.md https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "tirex-2-gifteval-zs",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://arxiv.org/pdf/2106.13008v5 https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "tirex",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used",
  "sources": "https://arxiv.org/pdf/2505.23719v2 https://arxiv.org/pdf/2106.13008v5 https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "toto-1.0",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used (via GIFT-Eval Pretrain (GiftEvalPretrain))",
  "sources": "https://arxiv.org/pdf/2505.14766v2 https://huggingface.co/Datadog/Toto-Open-Base-1.0/raw/main/README.md https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "toto-2.0-2.5b-ft",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used (via GIFT-Eval Pretrain (GiftEvalPretrain))",
  "sources": "https://huggingface.co/Datadog/Toto-2.0-2.5B-FT/raw/main/README.md https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "toto-2.0",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2605.20119v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "toto-2024",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used (via LOTSA (Large-scale Open Time Series Archive))",
  "sources": "https://arxiv.org/pdf/2407.07874v2 https://arxiv.org/pdf/2402.02592v2 https://arxiv.org/pdf/2106.13008v5 https://arxiv.org/pdf/2410.10393v2"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "ttm-r3",
  "level": "related",
  "reason": "related dataset CDC FluView ILINet (LOTSA) was used (via GIFT-Eval Pretrain (GiftEvalPretrain))",
  "sources": "https://huggingface.co/ibm-research/ttm-r3/raw/main/README.md https://arxiv.org/pdf/2410.10393v2 https://arxiv.org/pdf/2106.13008v5"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "units-few-shot-anomaly-detection",
  "level": "train_seen",
  "reason": "via base model UniTS: only the training split was used, not the benchmark's test data",
  "sources": "https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py"
 },
 {
  "benchmark": "long-horizon",
  "dataset": "ili",
  "model": "units",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py"
 },
 {
  "benchmark": "tabarena",
  "dataset": "airfoil-self-noise",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "airfoil-self-noise",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "airfoil-self-noise",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "airfoil-self-noise",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "airfoil-self-noise",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "airfoil-self-noise",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "airfoil-self-noise",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "airfoil-self-noise",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "airfoil-self-noise",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "airfoil-self-noise",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "airfoil-self-noise",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "airfoil-self-noise",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "airfoil-self-noise",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "airfoil-self-noise",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "airfoil-self-noise",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "airfoil-self-noise",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "airfoil-self-noise",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "amazon-employee-access",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "amazon-employee-access",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "amazon-employee-access",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "amazon-employee-access",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "amazon-employee-access",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "amazon-employee-access",
  "model": "real-tabpfn",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2507.03971v1"
 },
 {
  "benchmark": "tabarena",
  "dataset": "amazon-employee-access",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "amazon-employee-access",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "amazon-employee-access",
  "model": "tabdpt-v1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2410.18164v3 https://arxiv.org/pdf/2506.16791v4"
 },
 {
  "benchmark": "tabarena",
  "dataset": "amazon-employee-access",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "amazon-employee-access",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "amazon-employee-access",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "amazon-employee-access",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "amazon-employee-access",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "amazon-employee-access",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "amazon-employee-access",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "amazon-employee-access",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "anneal",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "anneal",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "anneal",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "anneal",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "anneal",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "anneal",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "anneal",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "anneal",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "anneal",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "anneal",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "anneal",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "anneal",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "anneal",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "anneal",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "anneal",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "anneal",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "anneal",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "another-dataset-on-used-fiat-500",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "another-dataset-on-used-fiat-500",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "another-dataset-on-used-fiat-500",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "another-dataset-on-used-fiat-500",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "another-dataset-on-used-fiat-500",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "another-dataset-on-used-fiat-500",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "another-dataset-on-used-fiat-500",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "another-dataset-on-used-fiat-500",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "another-dataset-on-used-fiat-500",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "another-dataset-on-used-fiat-500",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "another-dataset-on-used-fiat-500",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "another-dataset-on-used-fiat-500",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "another-dataset-on-used-fiat-500",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "another-dataset-on-used-fiat-500",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "another-dataset-on-used-fiat-500",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "another-dataset-on-used-fiat-500",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "another-dataset-on-used-fiat-500",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "apsfailure",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "apsfailure",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "apsfailure",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "apsfailure",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "apsfailure",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "apsfailure",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "apsfailure",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "apsfailure",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "apsfailure",
  "model": "tabdpt-v1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2410.18164v3 https://arxiv.org/pdf/2506.16791v4"
 },
 {
  "benchmark": "tabarena",
  "dataset": "apsfailure",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "apsfailure",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "apsfailure",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "apsfailure",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "apsfailure",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "apsfailure",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "apsfailure",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "apsfailure",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-marketing",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-marketing",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-marketing",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-marketing",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-marketing",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-marketing",
  "model": "real-tabpfn",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2507.03971v1"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-marketing",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-marketing",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-marketing",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-marketing",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-marketing",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-marketing",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-marketing",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-marketing",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-marketing",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-marketing",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-marketing",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-customer-churn",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-customer-churn",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-customer-churn",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-customer-churn",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-customer-churn",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-customer-churn",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-customer-churn",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-customer-churn",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-customer-churn",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-customer-churn",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-customer-churn",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-customer-churn",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-customer-churn",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-customer-churn",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-customer-churn",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-customer-churn",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bank-customer-churn",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bioresponse",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bioresponse",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bioresponse",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bioresponse",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bioresponse",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bioresponse",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "bioresponse",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bioresponse",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "bioresponse",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "bioresponse",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bioresponse",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bioresponse",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bioresponse",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bioresponse",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bioresponse",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bioresponse",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "bioresponse",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "blood-transfusion-service-center",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "blood-transfusion-service-center",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "blood-transfusion-service-center",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "blood-transfusion-service-center",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "blood-transfusion-service-center",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "blood-transfusion-service-center",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "blood-transfusion-service-center",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "blood-transfusion-service-center",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "blood-transfusion-service-center",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "blood-transfusion-service-center",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "blood-transfusion-service-center",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "blood-transfusion-service-center",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "blood-transfusion-service-center",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "blood-transfusion-service-center",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "blood-transfusion-service-center",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "blood-transfusion-service-center",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "blood-transfusion-service-center",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "churn",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "churn",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "churn",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "churn",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "churn",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "churn",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "churn",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "churn",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "churn",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "churn",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "churn",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "churn",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "churn",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "churn",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "churn",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "churn",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "churn",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "coil2000-insurance-policies",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "coil2000-insurance-policies",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "coil2000-insurance-policies",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "coil2000-insurance-policies",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "coil2000-insurance-policies",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "coil2000-insurance-policies",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "coil2000-insurance-policies",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "coil2000-insurance-policies",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "coil2000-insurance-policies",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "coil2000-insurance-policies",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "coil2000-insurance-policies",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "coil2000-insurance-policies",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "coil2000-insurance-policies",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "coil2000-insurance-policies",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "coil2000-insurance-policies",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "coil2000-insurance-policies",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "coil2000-insurance-policies",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "concrete-compressive-strength",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "concrete-compressive-strength",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "concrete-compressive-strength",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "concrete-compressive-strength",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "concrete-compressive-strength",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "concrete-compressive-strength",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "concrete-compressive-strength",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "concrete-compressive-strength",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "concrete-compressive-strength",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "concrete-compressive-strength",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "concrete-compressive-strength",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "concrete-compressive-strength",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "concrete-compressive-strength",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "concrete-compressive-strength",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "concrete-compressive-strength",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "concrete-compressive-strength",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "concrete-compressive-strength",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-g",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-g",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-g",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-g",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-g",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-g",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-g",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-g",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-g",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-g",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-g",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-g",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-g",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-g",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-g",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-g",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-g",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-card-clients-default",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-card-clients-default",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-card-clients-default",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-card-clients-default",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-card-clients-default",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-card-clients-default",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-card-clients-default",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-card-clients-default",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-card-clients-default",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-card-clients-default",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-card-clients-default",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-card-clients-default",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-card-clients-default",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-card-clients-default",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-card-clients-default",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-card-clients-default",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "credit-card-clients-default",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "customer-satisfaction-in-airline",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "customer-satisfaction-in-airline",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "customer-satisfaction-in-airline",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "customer-satisfaction-in-airline",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "customer-satisfaction-in-airline",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "customer-satisfaction-in-airline",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "customer-satisfaction-in-airline",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "customer-satisfaction-in-airline",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "customer-satisfaction-in-airline",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "customer-satisfaction-in-airline",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "customer-satisfaction-in-airline",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "customer-satisfaction-in-airline",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "customer-satisfaction-in-airline",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "customer-satisfaction-in-airline",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "customer-satisfaction-in-airline",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "customer-satisfaction-in-airline",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "customer-satisfaction-in-airline",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes130us",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes130us",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes130us",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes130us",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes130us",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes130us",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes130us",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes130us",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes130us",
  "model": "tabdpt-v1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2410.18164v3 https://arxiv.org/pdf/2506.16791v4"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes130us",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes130us",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes130us",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes130us",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes130us",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes130us",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes130us",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diabetes130us",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diamonds",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diamonds",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diamonds",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diamonds",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diamonds",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diamonds",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "diamonds",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diamonds",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "diamonds",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "diamonds",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diamonds",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diamonds",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diamonds",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diamonds",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diamonds",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diamonds",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "diamonds",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "e-commere-shipping-data",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "e-commere-shipping-data",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "e-commere-shipping-data",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "e-commere-shipping-data",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "e-commere-shipping-data",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "e-commere-shipping-data",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "e-commere-shipping-data",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "e-commere-shipping-data",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "e-commere-shipping-data",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "e-commere-shipping-data",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "e-commere-shipping-data",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "e-commere-shipping-data",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "e-commere-shipping-data",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "e-commere-shipping-data",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "e-commere-shipping-data",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "e-commere-shipping-data",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "e-commere-shipping-data",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "fitness-club",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "fitness-club",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "fitness-club",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "fitness-club",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "fitness-club",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "fitness-club",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "fitness-club",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "fitness-club",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "fitness-club",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "fitness-club",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "fitness-club",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "fitness-club",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "fitness-club",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "fitness-club",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "fitness-club",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "fitness-club",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "fitness-club",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "food-delivery-time",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "food-delivery-time",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "food-delivery-time",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "food-delivery-time",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "food-delivery-time",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "food-delivery-time",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "food-delivery-time",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "food-delivery-time",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "food-delivery-time",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "food-delivery-time",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "food-delivery-time",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "food-delivery-time",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "food-delivery-time",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "food-delivery-time",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "food-delivery-time",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "food-delivery-time",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "food-delivery-time",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "give-me-some-credit",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "give-me-some-credit",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "give-me-some-credit",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "give-me-some-credit",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "give-me-some-credit",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "give-me-some-credit",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "give-me-some-credit",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "give-me-some-credit",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "give-me-some-credit",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "give-me-some-credit",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "give-me-some-credit",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "give-me-some-credit",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "give-me-some-credit",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "give-me-some-credit",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "give-me-some-credit",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "give-me-some-credit",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "give-me-some-credit",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hazelnut-spread-contaminant-detection",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hazelnut-spread-contaminant-detection",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hazelnut-spread-contaminant-detection",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hazelnut-spread-contaminant-detection",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hazelnut-spread-contaminant-detection",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hazelnut-spread-contaminant-detection",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "hazelnut-spread-contaminant-detection",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hazelnut-spread-contaminant-detection",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "hazelnut-spread-contaminant-detection",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "hazelnut-spread-contaminant-detection",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hazelnut-spread-contaminant-detection",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hazelnut-spread-contaminant-detection",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hazelnut-spread-contaminant-detection",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hazelnut-spread-contaminant-detection",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hazelnut-spread-contaminant-detection",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hazelnut-spread-contaminant-detection",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hazelnut-spread-contaminant-detection",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "healthcare-insurance-expenses",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "healthcare-insurance-expenses",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "healthcare-insurance-expenses",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "healthcare-insurance-expenses",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "healthcare-insurance-expenses",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "healthcare-insurance-expenses",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "healthcare-insurance-expenses",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "healthcare-insurance-expenses",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "healthcare-insurance-expenses",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "healthcare-insurance-expenses",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "healthcare-insurance-expenses",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "healthcare-insurance-expenses",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "healthcare-insurance-expenses",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "healthcare-insurance-expenses",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "healthcare-insurance-expenses",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "healthcare-insurance-expenses",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "healthcare-insurance-expenses",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "heloc",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "heloc",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "heloc",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "heloc",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "heloc",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "heloc",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "heloc",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "heloc",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "heloc",
  "model": "tabdpt-v1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2410.18164v3 https://arxiv.org/pdf/2506.16791v4"
 },
 {
  "benchmark": "tabarena",
  "dataset": "heloc",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "heloc",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "heloc",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "heloc",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "heloc",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "heloc",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "heloc",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "heloc",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hiva-agnostic",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hiva-agnostic",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hiva-agnostic",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hiva-agnostic",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hiva-agnostic",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hiva-agnostic",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "hiva-agnostic",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hiva-agnostic",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "hiva-agnostic",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "hiva-agnostic",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hiva-agnostic",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hiva-agnostic",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hiva-agnostic",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hiva-agnostic",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hiva-agnostic",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hiva-agnostic",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hiva-agnostic",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "houses",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "houses",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "houses",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "houses",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "houses",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "houses",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "houses",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "houses",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "houses",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "houses",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "houses",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "houses",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "houses",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "houses",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "houses",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "houses",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "houses",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hr-analytics-job-change-of-data-scientists",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hr-analytics-job-change-of-data-scientists",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hr-analytics-job-change-of-data-scientists",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hr-analytics-job-change-of-data-scientists",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hr-analytics-job-change-of-data-scientists",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hr-analytics-job-change-of-data-scientists",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "hr-analytics-job-change-of-data-scientists",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hr-analytics-job-change-of-data-scientists",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "hr-analytics-job-change-of-data-scientists",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "hr-analytics-job-change-of-data-scientists",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hr-analytics-job-change-of-data-scientists",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hr-analytics-job-change-of-data-scientists",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hr-analytics-job-change-of-data-scientists",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hr-analytics-job-change-of-data-scientists",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hr-analytics-job-change-of-data-scientists",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hr-analytics-job-change-of-data-scientists",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "hr-analytics-job-change-of-data-scientists",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "in-vehicle-coupon-recommendation",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "in-vehicle-coupon-recommendation",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "in-vehicle-coupon-recommendation",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "in-vehicle-coupon-recommendation",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "in-vehicle-coupon-recommendation",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "in-vehicle-coupon-recommendation",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "in-vehicle-coupon-recommendation",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "in-vehicle-coupon-recommendation",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "in-vehicle-coupon-recommendation",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "in-vehicle-coupon-recommendation",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "in-vehicle-coupon-recommendation",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "in-vehicle-coupon-recommendation",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "in-vehicle-coupon-recommendation",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "in-vehicle-coupon-recommendation",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "in-vehicle-coupon-recommendation",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "in-vehicle-coupon-recommendation",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "in-vehicle-coupon-recommendation",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "is-this-a-good-customer",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "is-this-a-good-customer",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "is-this-a-good-customer",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "is-this-a-good-customer",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "is-this-a-good-customer",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "is-this-a-good-customer",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "is-this-a-good-customer",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "is-this-a-good-customer",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "is-this-a-good-customer",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "is-this-a-good-customer",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "is-this-a-good-customer",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "is-this-a-good-customer",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "is-this-a-good-customer",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "is-this-a-good-customer",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "is-this-a-good-customer",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "is-this-a-good-customer",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "is-this-a-good-customer",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "kddcup09-appetency",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "kddcup09-appetency",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "kddcup09-appetency",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "kddcup09-appetency",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "kddcup09-appetency",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "kddcup09-appetency",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "kddcup09-appetency",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "kddcup09-appetency",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "kddcup09-appetency",
  "model": "tabdpt-v1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2410.18164v3 https://arxiv.org/pdf/2506.16791v4"
 },
 {
  "benchmark": "tabarena",
  "dataset": "kddcup09-appetency",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "kddcup09-appetency",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "kddcup09-appetency",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "kddcup09-appetency",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "kddcup09-appetency",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "kddcup09-appetency",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "kddcup09-appetency",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "kddcup09-appetency",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "marketing-campaign",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "marketing-campaign",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "marketing-campaign",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "marketing-campaign",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "marketing-campaign",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "marketing-campaign",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "marketing-campaign",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "marketing-campaign",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "marketing-campaign",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "marketing-campaign",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "marketing-campaign",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "marketing-campaign",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "marketing-campaign",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "marketing-campaign",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "marketing-campaign",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "marketing-campaign",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "marketing-campaign",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "maternal-health-risk",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "maternal-health-risk",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "maternal-health-risk",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "maternal-health-risk",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "maternal-health-risk",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "maternal-health-risk",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "maternal-health-risk",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "maternal-health-risk",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "maternal-health-risk",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "maternal-health-risk",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "maternal-health-risk",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "maternal-health-risk",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "maternal-health-risk",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "maternal-health-risk",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "maternal-health-risk",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "maternal-health-risk",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "maternal-health-risk",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "miami-housing",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "miami-housing",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "miami-housing",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "miami-housing",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "miami-housing",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "miami-housing",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "miami-housing",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "miami-housing",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "miami-housing",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "miami-housing",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "miami-housing",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "miami-housing",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "miami-housing",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "miami-housing",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "miami-housing",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "miami-housing",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "miami-housing",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "mic",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "mic",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "mic",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "mic",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "mic",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "mic",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "mic",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "mic",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "mic",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "mic",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "mic",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "mic",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "mic",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "mic",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "mic",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "mic",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "mic",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "naticusdroid",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "naticusdroid",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "naticusdroid",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "naticusdroid",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "naticusdroid",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "naticusdroid",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "naticusdroid",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "naticusdroid",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "naticusdroid",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "naticusdroid",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "naticusdroid",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "naticusdroid",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "naticusdroid",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "naticusdroid",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "naticusdroid",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "naticusdroid",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "naticusdroid",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "online-shoppers-intention",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "online-shoppers-intention",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "online-shoppers-intention",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "online-shoppers-intention",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "online-shoppers-intention",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "online-shoppers-intention",
  "model": "real-tabpfn",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2507.03971v1"
 },
 {
  "benchmark": "tabarena",
  "dataset": "online-shoppers-intention",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "online-shoppers-intention",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "online-shoppers-intention",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "online-shoppers-intention",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "online-shoppers-intention",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "online-shoppers-intention",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "online-shoppers-intention",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "online-shoppers-intention",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "online-shoppers-intention",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "online-shoppers-intention",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "online-shoppers-intention",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "physiochemical-protein",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "physiochemical-protein",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "physiochemical-protein",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "physiochemical-protein",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "physiochemical-protein",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "physiochemical-protein",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "physiochemical-protein",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "physiochemical-protein",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "physiochemical-protein",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "physiochemical-protein",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "physiochemical-protein",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "physiochemical-protein",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "physiochemical-protein",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "physiochemical-protein",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "physiochemical-protein",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "physiochemical-protein",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "physiochemical-protein",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "polish-companies-bankruptcy",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "polish-companies-bankruptcy",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "polish-companies-bankruptcy",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "polish-companies-bankruptcy",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "polish-companies-bankruptcy",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "polish-companies-bankruptcy",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "polish-companies-bankruptcy",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "polish-companies-bankruptcy",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "polish-companies-bankruptcy",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "polish-companies-bankruptcy",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "polish-companies-bankruptcy",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "polish-companies-bankruptcy",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "polish-companies-bankruptcy",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "polish-companies-bankruptcy",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "polish-companies-bankruptcy",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "polish-companies-bankruptcy",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "polish-companies-bankruptcy",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-biodeg",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-biodeg",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-biodeg",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-biodeg",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-biodeg",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-biodeg",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-biodeg",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-biodeg",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-biodeg",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-biodeg",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-biodeg",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-biodeg",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-biodeg",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-biodeg",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-biodeg",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-biodeg",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-biodeg",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-tid-11",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-tid-11",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-tid-11",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-tid-11",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-tid-11",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-tid-11",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-tid-11",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-tid-11",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-tid-11",
  "model": "tabdpt-v1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2410.18164v3 https://arxiv.org/pdf/2506.16791v4"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-tid-11",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-tid-11",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-tid-11",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-tid-11",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-tid-11",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-tid-11",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-tid-11",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-tid-11",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-fish-toxicity",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-fish-toxicity",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-fish-toxicity",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-fish-toxicity",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-fish-toxicity",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-fish-toxicity",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-fish-toxicity",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-fish-toxicity",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-fish-toxicity",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-fish-toxicity",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-fish-toxicity",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-fish-toxicity",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-fish-toxicity",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-fish-toxicity",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-fish-toxicity",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-fish-toxicity",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "qsar-fish-toxicity",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "sdss17",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "sdss17",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "sdss17",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "sdss17",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "sdss17",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "sdss17",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "sdss17",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "sdss17",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "sdss17",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "sdss17",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "sdss17",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "sdss17",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "sdss17",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "sdss17",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "sdss17",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "sdss17",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "sdss17",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "seismic-bumps",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "seismic-bumps",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "seismic-bumps",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "seismic-bumps",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "seismic-bumps",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "seismic-bumps",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "seismic-bumps",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "seismic-bumps",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "seismic-bumps",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "seismic-bumps",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "seismic-bumps",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "seismic-bumps",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "seismic-bumps",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "seismic-bumps",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "seismic-bumps",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "seismic-bumps",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "seismic-bumps",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "splice",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "splice",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "splice",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "splice",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "splice",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "splice",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "splice",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "splice",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "splice",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "splice",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "splice",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "splice",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "splice",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "splice",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "splice",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "splice",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "splice",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "students-dropout-and-academic-success",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "students-dropout-and-academic-success",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "students-dropout-and-academic-success",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "students-dropout-and-academic-success",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "students-dropout-and-academic-success",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "students-dropout-and-academic-success",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "students-dropout-and-academic-success",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "students-dropout-and-academic-success",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "students-dropout-and-academic-success",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "students-dropout-and-academic-success",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "students-dropout-and-academic-success",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "students-dropout-and-academic-success",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "students-dropout-and-academic-success",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "students-dropout-and-academic-success",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "students-dropout-and-academic-success",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "students-dropout-and-academic-success",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "students-dropout-and-academic-success",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "superconductivity",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "superconductivity",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "superconductivity",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "superconductivity",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "superconductivity",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "superconductivity",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "superconductivity",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "superconductivity",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "superconductivity",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "superconductivity",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "superconductivity",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "superconductivity",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "superconductivity",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "superconductivity",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "superconductivity",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "superconductivity",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "superconductivity",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "taiwanese-bankruptcy-prediction",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "taiwanese-bankruptcy-prediction",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "taiwanese-bankruptcy-prediction",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "taiwanese-bankruptcy-prediction",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "taiwanese-bankruptcy-prediction",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "taiwanese-bankruptcy-prediction",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "taiwanese-bankruptcy-prediction",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "taiwanese-bankruptcy-prediction",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "taiwanese-bankruptcy-prediction",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "taiwanese-bankruptcy-prediction",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "taiwanese-bankruptcy-prediction",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "taiwanese-bankruptcy-prediction",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "taiwanese-bankruptcy-prediction",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "taiwanese-bankruptcy-prediction",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "taiwanese-bankruptcy-prediction",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "taiwanese-bankruptcy-prediction",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "taiwanese-bankruptcy-prediction",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "website-phishing",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "website-phishing",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "website-phishing",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "website-phishing",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "website-phishing",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "website-phishing",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "website-phishing",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "website-phishing",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "website-phishing",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "website-phishing",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "website-phishing",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "website-phishing",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "website-phishing",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "website-phishing",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "website-phishing",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "website-phishing",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "website-phishing",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "wine-quality",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "wine-quality",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "wine-quality",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "wine-quality",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "wine-quality",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "wine-quality",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "wine-quality",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "wine-quality",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "wine-quality",
  "model": "tabdpt-v1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part",
  "sources": "https://arxiv.org/pdf/2410.18164v3 https://arxiv.org/pdf/2506.16791v4"
 },
 {
  "benchmark": "tabarena",
  "dataset": "wine-quality",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "wine-quality",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "wine-quality",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "wine-quality",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "wine-quality",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "wine-quality",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "wine-quality",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "wine-quality",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "tabarena",
  "dataset": "jm1",
  "model": "limix-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.17488v1 https://huggingface.co/stable-ai/LimiX-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "jm1",
  "model": "limix-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.03505v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "jm1",
  "model": "mitra-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/mitra-classifier/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "jm1",
  "model": "mitra-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2609.04540v1 https://huggingface.co/autogluon/mitra-classifier-2/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "jm1",
  "model": "real-tabpfn-2-5",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2511.08667v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "jm1",
  "model": "real-tabpfn",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "jm1",
  "model": "tabdpt-v1-2",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2608.01400v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "jm1",
  "model": "tabdpt-v1-3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "jm1",
  "model": "tabdpt-v1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "tabarena",
  "dataset": "jm1",
  "model": "tabicl-v1-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/soda-inria/tabicl/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "jm1",
  "model": "tabicl-v1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2502.05564v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "jm1",
  "model": "tabicl-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2602.11139v2"
 },
 {
  "benchmark": "tabarena",
  "dataset": "jm1",
  "model": "tabpfn-2-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2511.08667v2 https://huggingface.co/Prior-Labs/tabpfn_2_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "jm1",
  "model": "tabpfn-2-6",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_2_6/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "jm1",
  "model": "tabpfn-3-5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/Prior-Labs/tabpfn_3_5/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "jm1",
  "model": "tabpfn-3",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2605.13986v2 https://huggingface.co/Prior-Labs/tabpfn_3/raw/main/README.md"
 },
 {
  "benchmark": "tabarena",
  "dataset": "jm1",
  "model": "tabpfn-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://www.nature.com/articles/s41586-024-08328-6"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "dada",
  "level": "test_seen",
  "reason": "a contamination analysis found test examples in the pre-training data (Third-party (TimeRCD authors): DADA's pre-training corpus (IOPS, MGAB, YAHOO) heavily overlaps their zero-shot test benchmarks drawn from TSB-AD.) [third-party study]",
  "sources": "https://arxiv.org/pdf/2509.21190v5"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "moirai-2.0-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "moment-1",
  "level": "related",
  "reason": "related dataset UCR Time Series Anomaly Archive (Hexagon ML/UCR, KDD Cup 2021) was used (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE https://raw.githubusercontent.com/moment-timeseries-foundation-model/moment/main/momentfm/data/anomaly_detection_dataset.py https://proceedings.neurips.cc/paper_files/paper/2024/file/c3f3c690b7a99fba16d0efd35cb83b2c-Paper-Datasets_and_Benchmarks_Track.pdf https://raw.githubusercontent.com/TheDatumOrg/TSB-AD/main/Datasets/File_List/TSB-AD-U.csv"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "sundial",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "timer-s1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/gifteval/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "toto-1.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "toto-2.0-2.5b-ft",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "toto-2.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.13033v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "units-few-shot-anomaly-detection",
  "level": "related",
  "reason": "in fine-tuning data: related dataset Server Machine Dataset (SMD) was used",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh https://raw.githubusercontent.com/mims-harvard/UniTS/main/data_provider/data_loader.py https://proceedings.neurips.cc/paper_files/paper/2024/file/c3f3c690b7a99fba16d0efd35cb83b2c-Paper-Datasets_and_Benchmarks_Track.pdf https://raw.githubusercontent.com/TheDatumOrg/TSB-AD/main/Datasets/File_List/TSB-AD-U.csv"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-u",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "moirai-2.0-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "moment-1",
  "level": "related",
  "reason": "related dataset Server Machine Dataset (SMD) was used (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE https://arxiv.org/pdf/2509.21190v5 https://proceedings.neurips.cc/paper_files/paper/2024/file/c3f3c690b7a99fba16d0efd35cb83b2c-Paper-Datasets_and_Benchmarks_Track.pdf https://raw.githubusercontent.com/TheDatumOrg/TSB-AD/main/Datasets/File_List/TSB-AD-M.csv"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "sundial",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "timer-s1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/gifteval/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "toto-1.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "toto-2.0-2.5b-ft",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "toto-2.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.13033v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "units-few-shot-anomaly-detection",
  "level": "related",
  "reason": "in fine-tuning data: related dataset Server Machine Dataset (SMD) was used",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh https://raw.githubusercontent.com/mims-harvard/UniTS/main/data_provider/data_loader.py https://proceedings.neurips.cc/paper_files/paper/2024/file/c3f3c690b7a99fba16d0efd35cb83b2c-Paper-Datasets_and_Benchmarks_Track.pdf https://raw.githubusercontent.com/TheDatumOrg/TSB-AD/main/Datasets/File_List/TSB-AD-M.csv"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "tsb-ad-m",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "dada",
  "level": "related",
  "reason": "a contamination analysis found test examples of the related dataset TSB-AD-U (TSB-AD univariate) (Third-party (TimeRCD authors): DADA's pre-training corpus (IOPS, MGAB, YAHOO) heavily overlaps their zero-shot test benchmarks drawn from TSB-AD.) [third-party study]; the authors state this dataset was excluded",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://proceedings.neurips.cc/paper_files/paper/2024/file/c3f3c690b7a99fba16d0efd35cb83b2c-Paper-Datasets_and_Benchmarks_Track.pdf https://raw.githubusercontent.com/TheDatumOrg/TSB-AD/main/Datasets/File_List/TSB-AD-U.csv https://arxiv.org/pdf/2405.15273v4"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "moirai-2.0-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "moment-1",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE https://raw.githubusercontent.com/moment-timeseries-foundation-model/moment/main/momentfm/data/anomaly_detection_dataset.py"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "sundial",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "timer-s1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "timeradar",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2602.19068v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/gifteval/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "toto-1.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "toto-2.0-2.5b-ft",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "toto-2.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "ucr-anomaly-archive",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "dada",
  "level": "related",
  "reason": "a contamination analysis found test examples of the related dataset TSB-AD-U (TSB-AD univariate) (Third-party (TimeRCD authors): DADA's pre-training corpus (IOPS, MGAB, YAHOO) heavily overlaps their zero-shot test benchmarks drawn from TSB-AD.) [third-party study]; the authors state this dataset was excluded",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://proceedings.neurips.cc/paper_files/paper/2024/file/c3f3c690b7a99fba16d0efd35cb83b2c-Paper-Datasets_and_Benchmarks_Track.pdf https://raw.githubusercontent.com/TheDatumOrg/TSB-AD/main/Datasets/File_List/TSB-AD-U.csv https://arxiv.org/pdf/2405.15273v4"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "moirai-2.0-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "moment-1",
  "level": "test_seen",
  "reason": "its test split was used for training (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE https://arxiv.org/pdf/2509.21190v5"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "sundial",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "timer-s1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "timeradar",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2602.19068v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/gifteval/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "toto-1.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "toto-2.0-2.5b-ft",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "toto-2.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "units-few-shot-anomaly-detection",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh https://raw.githubusercontent.com/mims-harvard/UniTS/main/data_provider/data_loader.py"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smd",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "dada",
  "level": "related",
  "reason": "a contamination analysis found test examples of the related dataset TSB-AD-U (TSB-AD univariate) (Third-party (TimeRCD authors): DADA's pre-training corpus (IOPS, MGAB, YAHOO) heavily overlaps their zero-shot test benchmarks drawn from TSB-AD.) [third-party study]; the authors state this dataset was excluded",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://proceedings.neurips.cc/paper_files/paper/2024/file/c3f3c690b7a99fba16d0efd35cb83b2c-Paper-Datasets_and_Benchmarks_Track.pdf https://raw.githubusercontent.com/TheDatumOrg/TSB-AD/main/Datasets/File_List/TSB-AD-U.csv https://arxiv.org/pdf/2405.15273v4"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "moirai-2.0-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "moment-1",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE https://arxiv.org/pdf/2509.21190v5"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "sundial",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "timer-s1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "timeradar",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2602.19068v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/gifteval/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "toto-1.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "toto-2.0-2.5b-ft",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "toto-2.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "units-few-shot-anomaly-detection",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "msl",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "dada",
  "level": "related",
  "reason": "a contamination analysis found test examples of the related dataset TSB-AD-U (TSB-AD univariate) (Third-party (TimeRCD authors): DADA's pre-training corpus (IOPS, MGAB, YAHOO) heavily overlaps their zero-shot test benchmarks drawn from TSB-AD.) [third-party study]; the authors state this dataset was excluded",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://proceedings.neurips.cc/paper_files/paper/2024/file/c3f3c690b7a99fba16d0efd35cb83b2c-Paper-Datasets_and_Benchmarks_Track.pdf https://raw.githubusercontent.com/TheDatumOrg/TSB-AD/main/Datasets/File_List/TSB-AD-U.csv https://arxiv.org/pdf/2405.15273v4"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "moirai-2.0-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "moment-1",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE https://arxiv.org/pdf/2509.21190v5"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "sundial",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "timer-s1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "timeradar",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2602.19068v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/gifteval/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "toto-1.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "toto-2.0-2.5b-ft",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "toto-2.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "units-few-shot-anomaly-detection",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "smap",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "dada",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2405.15273v4"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "moirai-2.0-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "moment-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "sundial",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "timer-s1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "timeradar",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2602.19068v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/gifteval/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "toto-1.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "toto-2.0-2.5b-ft",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "toto-2.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "units-few-shot-anomaly-detection",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "psm",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "dada",
  "level": "related",
  "reason": "a contamination analysis found test examples of the related dataset TSB-AD-U (TSB-AD univariate) (Third-party (TimeRCD authors): DADA's pre-training corpus (IOPS, MGAB, YAHOO) heavily overlaps their zero-shot test benchmarks drawn from TSB-AD.) [third-party study]; the authors state this dataset was excluded",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://proceedings.neurips.cc/paper_files/paper/2024/file/c3f3c690b7a99fba16d0efd35cb83b2c-Paper-Datasets_and_Benchmarks_Track.pdf https://raw.githubusercontent.com/TheDatumOrg/TSB-AD/main/Datasets/File_List/TSB-AD-U.csv https://arxiv.org/pdf/2405.15273v4"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "moirai-2.0-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "moment-1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "sundial",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "timer-s1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "timeradar",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2602.19068v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/gifteval/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "toto-1.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "toto-2.0-2.5b-ft",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "toto-2.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "units-few-shot-anomaly-detection",
  "level": "train_seen",
  "reason": "in fine-tuning data: only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "swat",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "dada",
  "level": "test_seen",
  "reason": "the whole dataset was used",
  "sources": "https://arxiv.org/pdf/2405.15273v4 https://arxiv.org/pdf/2509.21190v5"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "moirai-2.0-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "moment-1",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE https://arxiv.org/pdf/2509.21190v5"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "sundial",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "timer-s1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/gifteval/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "toto-1.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "toto-2.0-2.5b-ft",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "toto-2.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "yahoo-s5",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "dada",
  "level": "related",
  "reason": "a contamination analysis found test examples of the related dataset TSB-AD-U (TSB-AD univariate) (Third-party (TimeRCD authors): DADA's pre-training corpus (IOPS, MGAB, YAHOO) heavily overlaps their zero-shot test benchmarks drawn from TSB-AD.) [third-party study]",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://proceedings.neurips.cc/paper_files/paper/2024/file/c3f3c690b7a99fba16d0efd35cb83b2c-Paper-Datasets_and_Benchmarks_Track.pdf https://raw.githubusercontent.com/TheDatumOrg/TSB-AD/main/Datasets/File_List/TSB-AD-U.csv"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "mantis-8m",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "moirai-2.0-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "moment-1",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE https://arxiv.org/pdf/2509.21190v5"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "nutime",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "sundial",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "timer-s1",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "timer-xl",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "timer",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/gifteval/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "toto-1.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "toto-2.0-2.5b-ft",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "toto-2.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "units-few-shot-anomaly-detection",
  "level": "not_seen",
  "reason": "absent from the complete list of training data and from its base models' data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh"
 },
 {
  "benchmark": "ts-anomaly",
  "dataset": "nab",
  "model": "units",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.00131v3"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "mantis-8m",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2502.15637v1 https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "moirai-2.0-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "moment-1",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "nutime",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "sundial",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))",
  "sources": "https://huggingface.co/thuml/sundial-base-128m/raw/main/README.md https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "timer-s1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))",
  "sources": "https://huggingface.co/thuml/Timer-S1/raw/main/README.md https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/gifteval/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "toto-1.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "toto-2.0-2.5b-ft",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "toto-2.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.13033v3 https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "units-few-shot-anomaly-detection",
  "level": "train_seen",
  "reason": "via base model UniTS: only the training split was used, not the benchmark's test data",
  "sources": "https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "ucr-archive",
  "model": "units",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "chronos-2-synth",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/autogluon/chronos-2-synth/raw/main/README.md https://arxiv.org/pdf/2510.15821v1"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "chronos-2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2510.15821v1 https://huggingface.co/amazon/chronos-2/raw/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "chronos-bolt",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "chronos-t5",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2403.07815v3 https://huggingface.co/datasets/autogluon/chronos_datasets/raw/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "dada",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "granite-ttm-r1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1/raw/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "granite-ttm-r2-1",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "granite-ttm-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "granite-ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "lag-llama",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2310.08278v3"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "mantis-8m",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2502.15637v1 https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "mantis-plus",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "mantis-v2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://raw.githubusercontent.com/vfeofanov/mantis/main/README.md https://arxiv.org/pdf/2602.17868v1"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "moirai-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2402.02592v2 https://huggingface.co/Salesforce/moirai-1.0-R-small/raw/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "moirai-1.1-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "moirai-2.0-r",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "moirai-moe-1.0-r",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2410.10469v1"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "moment-1",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data (via Time Series Pile)",
  "sources": "https://arxiv.org/pdf/2402.03885v3 https://huggingface.co/AutonLab/MOMENT-1-large/raw/main/README.md https://huggingface.co/api/datasets/AutonLab/Timeseries-PILE"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "nutime",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data; the test portion was explicitly excluded",
  "sources": "https://arxiv.org/pdf/2310.07402v3"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "sundial",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))",
  "sources": "https://huggingface.co/thuml/sundial-base-128m/raw/main/README.md https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "time-moe",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "time-rcd",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2509.21190v5 https://raw.githubusercontent.com/thu-sail-lab/Time-RCD/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "timer-s1",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))",
  "sources": "https://huggingface.co/thuml/Timer-S1/raw/main/README.md https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "timer-xl",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))",
  "sources": "https://arxiv.org/pdf/2410.04803v4 https://huggingface.co/thuml/timer-base-84m/raw/main/README.md https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md https://arxiv.org/pdf/2402.02368v3"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "timer",
  "level": "seen_split_unclear",
  "reason": "the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))",
  "sources": "https://arxiv.org/pdf/2402.02368v3 https://huggingface.co/datasets/thuml/UTSD/raw/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "timeradar",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "timesfm-1-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "timesfm-2-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "timesfm-2-5",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "timesfm-3-0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "tirex-1.1-gifteval",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "tirex-2-fevbench",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/fevbench/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "tirex-2-gifteval-pretrain",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "tirex-2-gifteval-zs",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2607.01204v1 https://raw.githubusercontent.com/NX-AI/tirex-2/main/examples/gifteval/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "tirex-2",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "tirex",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2505.23719v2"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "toto-1.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "toto-2.0-2.5b-ft",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "toto-2.0",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "toto-2024",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "tspulse-r1",
  "level": "not_seen",
  "reason": "the authors state the dataset was excluded from training",
  "sources": "https://arxiv.org/pdf/2505.13033v3"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "ttm-r3",
  "level": "unknown",
  "reason": "training data not disclosed in enough detail",
  "sources": ""
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "ttm-research-r2",
  "level": "not_seen",
  "reason": "absent from the complete list of training data",
  "sources": "https://arxiv.org/pdf/2401.03955v8 https://huggingface.co/ibm-research/ttm-research-r2/raw/main/README.md"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "units-few-shot-anomaly-detection",
  "level": "train_seen",
  "reason": "via base model UniTS: only the training split was used, not the benchmark's test data",
  "sources": "https://raw.githubusercontent.com/mims-harvard/UniTS/main/scripts/few_shot_anomaly_detection/UniTS_finetune_few_shot_anomaly_detection.sh https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py"
 },
 {
  "benchmark": "ts-classification",
  "dataset": "uea-archive",
  "model": "units",
  "level": "train_seen",
  "reason": "only the training split was used, not the benchmark's test data",
  "sources": "https://arxiv.org/pdf/2403.00131v3 https://raw.githubusercontent.com/mims-harvard/UniTS/main/exp/exp_pretrain.py"
 }
]