| { | |
| "checked_on": "2026-09-20", | |
| "all_training_tasks_covered": true, | |
| "all_rights_cleared": false, | |
| "records": [ | |
| { | |
| "task": "winogrande", | |
| "name": "WinoGrande", | |
| "attribution": "Sakaguchi et al.; Allen Institute for AI", | |
| "source_terms": "Apache-2.0", | |
| "source_url": "https://github.com/allenai/winogrande/blob/master/LICENSE", | |
| "citation_url": "https://arxiv.org/abs/1907.10641", | |
| "notice_file": "NOTICE-WINOGRANDE.txt", | |
| "requirements_and_limits": "Preserve license and applicable copyright/NOTICE notices; identify changes.", | |
| "main_examples": 20000, | |
| "repair_examples": 600 | |
| }, | |
| { | |
| "task": "xnli_de", | |
| "name": "XNLI (German)", | |
| "attribution": "Conneau et al.; Facebook Research", | |
| "source_terms": "CC-BY-NC-4.0", | |
| "source_url": "https://github.com/facebookresearch/XNLI/blob/main/LICENSE", | |
| "citation_url": "https://aclanthology.org/D18-1269/", | |
| "notice_file": "NOTICE-XNLI.txt", | |
| "requirements_and_limits": "Attribution, license link, change notice; noncommercial restriction. Training translations derive from MultiNLI.", | |
| "main_examples": 13000, | |
| "repair_examples": 250 | |
| }, | |
| { | |
| "task": "anli", | |
| "name": "ANLI", | |
| "attribution": "Nie et al.; Facebook Research", | |
| "source_terms": "CC-BY-NC-4.0", | |
| "source_url": "https://github.com/facebookresearch/anli/blob/main/LICENSE", | |
| "citation_url": "https://arxiv.org/abs/1910.14599", | |
| "notice_file": "NOTICE-ANLI.txt", | |
| "requirements_and_limits": "Attribution, license link, change notice; noncommercial restriction.", | |
| "main_examples": 10000, | |
| "repair_examples": 500 | |
| }, | |
| { | |
| "task": "massive_de", | |
| "name": "MASSIVE (German)", | |
| "attribution": "FitzGerald et al.; Amazon.com Inc. or its affiliates", | |
| "source_terms": "CC-BY-4.0", | |
| "source_url": "https://huggingface.co/datasets/AmazonScience/massive", | |
| "citation_url": "https://arxiv.org/abs/2204.08582", | |
| "notice_file": "NOTICE-MASSIVE.txt", | |
| "requirements_and_limits": "Retain Amazon attribution, license link and change notice.", | |
| "main_examples": 4000, | |
| "repair_examples": 500 | |
| }, | |
| { | |
| "task": "pawsx_de", | |
| "name": "PAWS-X (German)", | |
| "attribution": "Yang et al.; Google LLC", | |
| "source_terms": "Custom permissive dataset terms", | |
| "source_url": "https://github.com/google-research-datasets/paws/blob/master/LICENSE", | |
| "citation_url": "https://aclanthology.org/D19-1382/", | |
| "notice_file": "NOTICE-PAWS.txt", | |
| "requirements_and_limits": "Google source acknowledgement requested; preserve supplied disclaimer. Do not mislabel as Apache.", | |
| "main_examples": 6000, | |
| "repair_examples": 200 | |
| }, | |
| { | |
| "task": "hellaswag", | |
| "name": "HellaSwag", | |
| "attribution": "Zellers et al.", | |
| "source_terms": "MIT reported in pinned dataset card; upstream verification unavailable", | |
| "source_url": "https://huggingface.co/datasets/Rowan/hellaswag/blob/218ec52e09a7e7462a5400043bb9a69a41d06b76/README.md", | |
| "citation_url": "https://aclanthology.org/P19-1472/", | |
| "notice_file": null, | |
| "requirements_and_limits": "Original repository was unavailable during this audit. Original copyright notice and underlying source-text terms remain to be verified.", | |
| "main_examples": 8000, | |
| "repair_examples": 50 | |
| }, | |
| { | |
| "task": "piqa", | |
| "name": "PIQA", | |
| "attribution": "Bisk et al.", | |
| "source_terms": "AFL-3.0", | |
| "source_url": "https://yonatanbisk.com/piqa/", | |
| "citation_url": "https://arxiv.org/abs/1911.11641", | |
| "notice_file": "NOTICE-AFL-3.0.txt", | |
| "requirements_and_limits": "Original homepage explicitly names AFL-3.0, overriding unknown in the cached mirror card. Preserve attribution/change notices and applicable source/distribution obligations; see full terms.", | |
| "main_examples": 5000, | |
| "repair_examples": 0 | |
| }, | |
| { | |
| "task": "social_iqa", | |
| "name": "Social IQa", | |
| "attribution": "Sap et al.", | |
| "source_terms": "CC-BY-4.0 (pinned source card)", | |
| "source_url": "https://huggingface.co/datasets/allenai/social_i_qa/blob/537a2ec8ec565adc0b70b70752893e59e024df26/README.md", | |
| "citation_url": "https://aclanthology.org/D19-1454/", | |
| "notice_file": "NOTICE-CC-BY-4.0.txt", | |
| "requirements_and_limits": "Attribution, license link and change notice.", | |
| "main_examples": 5000, | |
| "repair_examples": 50 | |
| }, | |
| { | |
| "task": "boolq", | |
| "name": "BoolQ", | |
| "attribution": "Clark et al.; Google Research; Wikipedia contributors", | |
| "source_terms": "CC-BY-SA-3.0 (Google dataset card)", | |
| "source_url": "https://huggingface.co/datasets/google/boolq", | |
| "citation_url": "https://aclanthology.org/N19-1300/", | |
| "notice_file": "NOTICE-CC-BY-SA-3.0.txt", | |
| "requirements_and_limits": "Attribution, license link, modification notice and share-alike for covered adaptations; retain underlying Wikipedia source credits.", | |
| "main_examples": 1006, | |
| "repair_examples": 0 | |
| }, | |
| { | |
| "task": "sst2", | |
| "name": "Stanford Sentiment Treebank / SST-2 via GLUE", | |
| "attribution": "Socher et al.; Stanford NLP; GLUE maintainers", | |
| "source_terms": "Unresolved for original dataset in reviewed sources", | |
| "source_url": "https://nlp.stanford.edu/sentiment/", | |
| "citation_url": "https://aclanthology.org/D13-1170/", | |
| "notice_file": null, | |
| "requirements_and_limits": "GLUE explicitly refers to original task licenses. Do not apply the GLUE code license to SST-2 data.", | |
| "main_examples": 1000, | |
| "repair_examples": 0 | |
| }, | |
| { | |
| "task": "aokvqa", | |
| "name": "A-OKVQA via The Cauldron", | |
| "attribution": "Schwenk et al.; Allen Institute for AI", | |
| "source_terms": "Apache-2.0 repository; COCO images have separate rights", | |
| "source_url": "https://github.com/allenai/aokvqa/blob/main/LICENSE", | |
| "citation_url": "https://arxiv.org/abs/2206.01718", | |
| "notice_file": "NOTICE-AOKVQA.txt", | |
| "requirements_and_limits": "Preserve license/notices. This repository license does not establish permission for every underlying photograph.", | |
| "main_examples": 6000, | |
| "repair_examples": 100 | |
| }, | |
| { | |
| "task": "iconqa", | |
| "name": "IconQA via The Cauldron", | |
| "attribution": "Lu et al.", | |
| "source_terms": "CC-BY-NC-SA-4.0 (dataset)", | |
| "source_url": "https://github.com/lupantech/IconQA#license", | |
| "citation_url": "https://arxiv.org/abs/2110.13214", | |
| "notice_file": "NOTICE-CC-BY-NC-SA-4.0.txt", | |
| "requirements_and_limits": "Attribution, noncommercial use, change notice and share-alike for covered adaptations. Underlying icons/images may have additional terms.", | |
| "main_examples": 5000, | |
| "repair_examples": 100 | |
| }, | |
| { | |
| "task": "scienceqa", | |
| "name": "ScienceQA via The Cauldron", | |
| "attribution": "Lu et al.", | |
| "source_terms": "CC-BY-NC-SA-4.0 (dataset); code separately MIT", | |
| "source_url": "https://github.com/lupantech/ScienceQA#-licenses", | |
| "citation_url": "https://arxiv.org/abs/2209.09513", | |
| "notice_file": "NOTICE-CC-BY-NC-SA-4.0.txt", | |
| "requirements_and_limits": "Use dataset terms, not the code license. Attribution, noncommercial use, change notice and share-alike for covered adaptations.", | |
| "main_examples": 3000, | |
| "repair_examples": 100 | |
| }, | |
| { | |
| "task": "ai2d", | |
| "name": "AI2D via The Cauldron", | |
| "attribution": "Kembhavi et al.; Allen Institute for AI", | |
| "source_terms": "CC-BY-SA-4.0 (AI2-managed AWS registry entry)", | |
| "source_url": "https://registry.opendata.aws/allenai-diagrams/", | |
| "citation_url": "https://arxiv.org/abs/1603.07396", | |
| "notice_file": "NOTICE-CC-BY-SA-4.0.txt", | |
| "requirements_and_limits": "Attribution, license link, change notice and share-alike for covered adaptations. Data acquired through The Cauldron; original AI2 registry/terms accessed 2026-09-20. Individual diagram provenance not reconstructed.", | |
| "main_examples": 1994, | |
| "repair_examples": 100 | |
| }, | |
| { | |
| "task": "vsr", | |
| "name": "Visual Spatial Reasoning via The Cauldron", | |
| "attribution": "Liu, Emerson and Collier", | |
| "source_terms": "Apache-2.0 project; underlying COCO images separate", | |
| "source_url": "https://github.com/cambridgeltl/visual-spatial-reasoning", | |
| "citation_url": "https://aclanthology.org/2023.tacl-1.37/", | |
| "notice_file": "NOTICE-VSR.txt", | |
| "requirements_and_limits": "Preserve license/notices and changes; retain separate image-source obligations.", | |
| "main_examples": 1000, | |
| "repair_examples": 50 | |
| }, | |
| { | |
| "task": "textvqa_mc", | |
| "name": "TextVQA via The Cauldron", | |
| "attribution": "Singh et al.; Facebook AI Research", | |
| "source_terms": "CC-BY-4.0 dataset; OpenImages/Flickr image-source rights", | |
| "source_url": "https://huggingface.co/datasets/facebook/textvqa", | |
| "citation_url": "https://arxiv.org/abs/1904.08920", | |
| "notice_file": "NOTICE-CC-BY-4.0.txt", | |
| "requirements_and_limits": "Attribution, license link and modification notice. Original answers converted to choices with distractors; individual image-source attribution is not reconstructed here.", | |
| "main_examples": 2000, | |
| "repair_examples": 150 | |
| }, | |
| { | |
| "task": "docvqa_mc", | |
| "name": "DocVQA via The Cauldron", | |
| "attribution": "Mathew, Karatzas and Jawahar; UCSF Industry Documents Library (documents)", | |
| "source_terms": "Original challenge download terms; not verified", | |
| "source_url": "https://site.docvqa.org/datasets", | |
| "citation_url": "https://arxiv.org/abs/2007.00398", | |
| "notice_file": null, | |
| "requirements_and_limits": "Original site requires review of RRC download terms. Portal unavailable during audit; no blanket permissive license asserted. Individual document rights remain separate.", | |
| "main_examples": 2000, | |
| "repair_examples": 150 | |
| }, | |
| { | |
| "task": "corefud_en", | |
| "name": "GUM via CorefUD 1.4 / CRAC 2026", | |
| "attribution": "Zeldes; GUM annotators and original text authors; CorefUD team", | |
| "source_terms": "CC-BY-NC-SA-4.0 collection; constituent text licenses vary", | |
| "source_url": "https://github.com/UniversalDependencies/UD_English-GUM/blob/master/LICENSE.txt", | |
| "citation_url": "https://doi.org/10.1007/s10579-016-9343-x", | |
| "notice_file": "NOTICE-GUM.txt", | |
| "requirements_and_limits": "Credit original text sources and annotators. Selected document IDs, authors and source URLs are listed in coref_attributions.json. Document-level license verification remains incomplete.", | |
| "main_examples": 3368, | |
| "repair_examples": 400 | |
| }, | |
| { | |
| "task": "corefud_de", | |
| "name": "Potsdam Commentary Corpus via CorefUD 1.4 / CRAC 2026", | |
| "attribution": "Bourgonje and Stede; Potsdam corpus contributors; newspaper authors; CorefUD team", | |
| "source_terms": "CC-BY-NC-SA-4.0 (license linked by original corpus page)", | |
| "source_url": "https://angcl.ling.uni-potsdam.de/resources/pcc.html", | |
| "citation_url": "https://aclanthology.org/2020.lrec-1.133/", | |
| "notice_file": "NOTICE-CC-BY-NC-SA-4.0.txt", | |
| "requirements_and_limits": "Credit PCC and original newspaper sources; noncommercial/share-alike conditions. Preserve notices and identify transformations.", | |
| "main_examples": 632, | |
| "repair_examples": 200 | |
| }, | |
| { | |
| "task": "coref_0", | |
| "name": "Locally generated coref_0", | |
| "attribution": "This project", | |
| "source_terms": "Project-authored templates; no separate third-party dataset grant asserted", | |
| "source_url": null, | |
| "citation_url": null, | |
| "notice_file": null, | |
| "requirements_and_limits": "Computed labels; template-generated examples, not real conversations. Adapter license is separate.", | |
| "main_examples": 88, | |
| "repair_examples": 0 | |
| }, | |
| { | |
| "task": "coref_1", | |
| "name": "Locally generated coref_1", | |
| "attribution": "This project", | |
| "source_terms": "Project-authored templates; no separate third-party dataset grant asserted", | |
| "source_url": null, | |
| "citation_url": null, | |
| "notice_file": null, | |
| "requirements_and_limits": "Computed labels; template-generated examples, not real conversations. Adapter license is separate.", | |
| "main_examples": 176, | |
| "repair_examples": 0 | |
| }, | |
| { | |
| "task": "coref_2", | |
| "name": "Locally generated coref_2", | |
| "attribution": "This project", | |
| "source_terms": "Project-authored templates; no separate third-party dataset grant asserted", | |
| "source_url": null, | |
| "citation_url": null, | |
| "notice_file": null, | |
| "requirements_and_limits": "Computed labels; template-generated examples, not real conversations. Adapter license is separate.", | |
| "main_examples": 192, | |
| "repair_examples": 0 | |
| }, | |
| { | |
| "task": "coref_3", | |
| "name": "Locally generated coref_3", | |
| "attribution": "This project", | |
| "source_terms": "Project-authored templates; no separate third-party dataset grant asserted", | |
| "source_url": null, | |
| "citation_url": null, | |
| "notice_file": null, | |
| "requirements_and_limits": "Computed labels; template-generated examples, not real conversations. Adapter license is separate.", | |
| "main_examples": 160, | |
| "repair_examples": 0 | |
| }, | |
| { | |
| "task": "coref_4", | |
| "name": "Locally generated coref_4", | |
| "attribution": "This project", | |
| "source_terms": "Project-authored templates; no separate third-party dataset grant asserted", | |
| "source_url": null, | |
| "citation_url": null, | |
| "notice_file": null, | |
| "requirements_and_limits": "Computed labels; template-generated examples, not real conversations. Adapter license is separate.", | |
| "main_examples": 176, | |
| "repair_examples": 0 | |
| }, | |
| { | |
| "task": "coref_5", | |
| "name": "Locally generated coref_5", | |
| "attribution": "This project", | |
| "source_terms": "Project-authored templates; no separate third-party dataset grant asserted", | |
| "source_url": null, | |
| "citation_url": null, | |
| "notice_file": null, | |
| "requirements_and_limits": "Computed labels; template-generated examples, not real conversations. Adapter license is separate.", | |
| "main_examples": 208, | |
| "repair_examples": 0 | |
| }, | |
| { | |
| "task": "synthetic_dialogue", | |
| "name": "Locally generated synthetic_dialogue", | |
| "attribution": "This project", | |
| "source_terms": "Project-authored templates; no separate third-party dataset grant asserted", | |
| "source_url": null, | |
| "citation_url": null, | |
| "notice_file": null, | |
| "requirements_and_limits": "Computed labels; template-generated examples, not real conversations. Adapter license is separate.", | |
| "main_examples": 200, | |
| "repair_examples": 0 | |
| }, | |
| { | |
| "task": "synthetic_numeric", | |
| "name": "Locally generated synthetic_numeric", | |
| "attribution": "This project", | |
| "source_terms": "Project-authored templates; no separate third-party dataset grant asserted", | |
| "source_url": null, | |
| "citation_url": null, | |
| "notice_file": null, | |
| "requirements_and_limits": "Computed labels; template-generated examples, not real conversations. Adapter license is separate.", | |
| "main_examples": 100, | |
| "repair_examples": 1500 | |
| }, | |
| { | |
| "task": "synthetic_ocr", | |
| "name": "Locally generated synthetic_ocr", | |
| "attribution": "This project", | |
| "source_terms": "Project-authored templates; no separate third-party dataset grant asserted", | |
| "source_url": null, | |
| "citation_url": null, | |
| "notice_file": null, | |
| "requirements_and_limits": "Computed labels; template-generated examples, not real conversations. Adapter license is separate.", | |
| "main_examples": 500, | |
| "repair_examples": 0 | |
| }, | |
| { | |
| "task": "synthetic_rules", | |
| "name": "Locally generated synthetic_rules", | |
| "attribution": "This project", | |
| "source_terms": "Project-authored templates; no separate third-party dataset grant asserted", | |
| "source_url": null, | |
| "citation_url": null, | |
| "notice_file": null, | |
| "requirements_and_limits": "Computed labels; template-generated examples, not real conversations. Adapter license is separate.", | |
| "main_examples": 100, | |
| "repair_examples": 0 | |
| }, | |
| { | |
| "task": "synthetic_temporal", | |
| "name": "Locally generated synthetic_temporal", | |
| "attribution": "This project", | |
| "source_terms": "Project-authored templates; no separate third-party dataset grant asserted", | |
| "source_url": null, | |
| "citation_url": null, | |
| "notice_file": null, | |
| "requirements_and_limits": "Computed labels; template-generated examples, not real conversations. Adapter license is separate.", | |
| "main_examples": 100, | |
| "repair_examples": 1000 | |
| } | |
| ] | |
| } | |