Contamination Map Methodology

Vision / Model

OpenCLIP (LAION-2B checkpoints)

Organisation
LAION / OpenCLIP (mlfoundations)
Family
OpenCLIP
Released
2022-09-14 (HF repo creation of laion/CLIP-ViT-H-14-laion2B-s32B-b79K and laion/CLIP-ViT-B-32-laion2B-s34B-b79K (HF API createdAt))
Checkpoints
laion/CLIP-ViT-B-32-laion2B-s34B-b79K, laion/CLIP-ViT-B-16-laion2B-s34B-b88K, laion/CLIP-ViT-L-14-laion2B-s32B-b82K, laion/CLIP-ViT-H-14-laion2B-s32B-b79K, laion/CLIP-ViT-g-14-laion2B-s12B-b42K, laion/CLIP-ViT-g-14-laion2B-s34B-b88K, laion/CLIP-ViT-bigG-14-laion2B-39B-b160k
Weights
Open · MIT
Links
PaperModel cardCode

Training data disclosure: Full list

Level determined for 50% of benchmark cells.

“This model was trained with the 2 Billion sample English subset of LAION-5B”
huggingface.co · model card · Training Data · retrieved 2026-10-15

Training corpora

Benchmark cells

Industrial anomaly detection

Unknown MVTec AD · training data not disclosed in enough detail

Not seen MVTec AD 2 · its test items were published from 2025-03-27, after the model's stated training-data cutoff (2022-01-31)

Evidence (3)
“The data was filtered from September 2021 to January 2022, but those who created the sites might have included content from before then.”
arxiv.org · paper · Datasheet, Q26 · retrieved 2026-10-15
“arXiv:2503.21622v1 [cs.CV] 27 Mar 2025”
arxiv.org · paper · header · retrieved 2026-10-15
“All image data is available at https://www.mvtec.com/company/research/datasets/mvtec-ad-2.”
arxiv.org · paper · Abstract · retrieved 2026-10-15

Not seen MVTec LOCO AD · its test items were published from 2022-02-22, after the model's stated training-data cutoff (2022-01-31)

Evidence (2)
“The data was filtered from September 2021 to January 2022, but those who created the sites might have included content from before then.”
arxiv.org · paper · Datasheet, Q26 · retrieved 2026-10-15
“Published: 22 February 2022”
link.springer.com · paper · article metadata · retrieved 2026-10-15

Not seen VisA · its test items were published from 2022-07-28, after the model's stated training-data cutoff (2022-01-31)

Evidence (3)
“The data was filtered from September 2021 to January 2022, but those who created the sites might have included content from before then.”
arxiv.org · paper · Datasheet, Q26 · retrieved 2026-10-15
“arXiv:2207.14315v1 [cs.CV] 28 Jul 2022”
arxiv.org · paper · header · retrieved 2026-10-15
“We release the Visual Anomaly (VisA) Dataset consisting of 10,821 high-resolution color images (9,621 normal and 1,200 anomalous samples) covering 12 objects in 3 domains”
arxiv.org · paper · Abstract · retrieved 2026-10-15

Not seen Real-IAD · its test items were published from 2024-03-19, after the model's stated training-data cutoff (2022-01-31)

Evidence (2)
“The data was filtered from September 2021 to January 2022, but those who created the sites might have included content from before then.”
arxiv.org · paper · Datasheet, Q26 · retrieved 2026-10-15
“arXiv:2403.12580v1 [cs.CV] 19 Mar 2024”
arxiv.org · paper · header · retrieved 2026-10-15

Unknown MPDD · training data not disclosed in enough detail

Unknown BTAD · training data not disclosed in enough detail

Unknown KolektorSDD2 · training data not disclosed in enough detail

Notes

The list includes the ViT checkpoints most often used as AD backbones. pretrained.py also has ConvNeXt and RoBERTa-text LAION-2B checkpoints with the laion2b_* tags. disclosure is full_list because the training set is one published dataset, but that dataset is an open-ended web scrape. The scaling-law paper's pHash duplicate check (Appendix B.1) covered only ImageNet variants and VTAB and removed nothing from pre-training. training_cutoff is on the laion-2b corpus.

Sources