Contamination MapText / LLM / Model
gpt-oss
- Organisation
- OpenAI
- Family
- gpt-oss
- Released
- 2025-08-05 (Date on the gpt-oss model card (August 5, 2025), https://arxiv.org/pdf/2508.10925v1)
- Training cutoff
- 2024-06-01
Evidence
“Our model has a knowledge cutoff of June 2024.”
arxiv.org · paper · Section 2.4 (Pretraining) · retrieved 2026-10-15 - Checkpoints
- openai/gpt-oss-120b, openai/gpt-oss-20b
- Weights
- Open · Apache 2.0
- Links
- PaperModel cardCode
Training data disclosure: Vague
Level determined for 6% of benchmark cells.
- TestTest data seen0
- SeenSeen, split unclear0
- TrainTraining data seen0
- Related data seen0
- NoNot seen1
- ?Unknown14
“We train the models on a text-only dataset with trillions of tokens, with a focus on STEM, coding, and general knowledge.”
Other declared training data
-
text-only pre-training dataset (trillions of tokens; STEM, coding, general knowledge; CBRN-filtered) (not yet matched to a dataset) · split: unknown
Sources not stated (web data is not mentioned explicitly); open_ended left null.
Evidence
“We train the models on a text-only dataset with trillions of tokens, with a focus on STEM, coding, and general knowledge.”
arxiv.org · paper · Section 2.4 (Pretraining) · retrieved 2026-10-15 -
CoT RL training problems (coding, math, science) (not yet matched to a dataset) · split: unknown · fine-tuning
Evidence
“Our training dataset consists of a wide range of problems from coding, math, science, and more.”
arxiv.org · paper · Section 2.5 (Post-Training for Reasoning and Tool Use) · retrieved 2026-10-15
Benchmark cells
Classic LLM benchmarks
Unknown MMLU (Measuring Massive Multitask Language Understanding) · training data not disclosed in enough detail
Unknown GSM8K (Grade School Math 8K) · training data not disclosed in enough detail
Unknown MATH (Mathematics Aptitude Test of Heuristics) · training data not disclosed in enough detail
Unknown HumanEval · training data not disclosed in enough detail
Unknown ARC-Challenge (AI2 Reasoning Challenge, Challenge Set) · training data not disclosed in enough detail
Unknown HellaSwag · training data not disclosed in enough detail
Unknown TruthfulQA · training data not disclosed in enough detail
Unknown WinoGrande · training data not disclosed in enough detail
Recent LLM benchmarks
Unknown GPQA Diamond · training data not disclosed in enough detail
Unknown MMLU-Pro · training data not disclosed in enough detail
Unknown LiveCodeBench (code generation, release_v6) · training data not disclosed in enough detail
Unknown SWE-bench Verified · training data not disclosed in enough detail
Unknown AIME 2024 · training data not disclosed in enough detail
Not seen AIME 2025 · its test items were published from 2025-02-06, after the model's stated training-data cutoff (2024-06-01)
Evidence (3)
“Our model has a knowledge cutoff of June 2024.”
“2025 AIME I problems and solutions. The test was held on Thursday, February 6, 2025.”
“This year the AIME I will take place on February 6, 2025, while the alternate AIME II will take place on February 12, 2025.”
Unknown MATH-500 · training data not disclosed in enough detail
Notes
No decontamination or contamination analysis stated. gpt-oss-safeguard-120b/20b (Sept 2025, safety fine-tunes) not covered.
Sources
- gpt-oss-120b & gpt-oss-20b Model Card · paper · retrieved 2026-10-15
- gpt-oss-120b model card · model card · retrieved 2026-10-15
- gpt-oss-20b model card · model card · retrieved 2026-10-15