{
 "id": "geodesic-research/bipia",
 "owner": "geodesic-research",
 "name": "bipia",
 "downloads": 706,
 "likes": 0,
 "last_modified": "2026-08-11T11:45:42.000Z",
 "stated_licence": "cc-by-sa-4.0",
 "stated_from": "tag",
 "licence_name": null,
 "licence_link": null,
 "file_licence": null,
 "licence_files": [],
 "sources": [],
 "gated": false,
 "tasks": [],
 "citation": "This dataset assembles third-party benchmark **contexts** with prompt-injection **attack**\nstrings, following the [microsoft/BIPIA](https://github.com/microsoft/BIPIA) benchmark\n(arXiv:2312.14197; BIPIA code is MIT). The context data is redistributed under the licenses\nof its upstream sources, so this mirror is released under **CC BY-SA 4.0** (the most\nrestrictive of the included licenses):\n\n| Task | Context source | License |\n|---|---|---|\n| email | Invoices data (OpenAI Evals) | MIT |\n| table | WikiTableQuestions (Pasupat & Liang, 2015) | CC BY-SA 4.0 |\n| code  | Stack Overflow (100 questions) | CC BY-SA 4.0 |\n\n**Attribution.** Table contexts © the WikiTableQuestions authors (Pasupat & Liang, 2015);\ncode contexts © the respective Stack Overflow authors (see upstream question URLs); attack\ntaxonomy and assembly from microsoft/BIPIA. **Changes:** each context is crossed with BIPIA\ntest attacks at three insert positions (start / middle / end); context text is not otherwise\naltered. Assembly details are in the dataset-builder config.",
 "signal": "permissive",
 "reason": "The declared licence (cc-by-sa-4.0) is classified as share-alike.",
 "revision": "003f58372e1696726b7cddde274a74bee0b1938a",
 "stated_evidence": "https://huggingface.co/datasets/geodesic-research/bipia/blob/003f58372e1696726b7cddde274a74bee0b1938a/README.md",
 "file_evidence": "https://huggingface.co/datasets/geodesic-research/bipia/tree/003f58372e1696726b7cddde274a74bee0b1938a",
 "card_nc": false
}