MedConform-Bench: experimental data and code for medical LLM conformity evaluation
收藏官方服务：
资源简介：
De-identified experimental records and analysis code for MedConform-Bench (PLOS ONE submission). Includes:- Main study: 13,500 model response records (6 models × 3 datasets × 5 protocols × 150 items × 3 repeats)- Mitigation study: 8,100 Wrong-Guidance records (CP-1/CP-2/baseline)- Aggregated summary tables, high-risk dual-review annotations (153 items), and reproduction scripts Public benchmark items are from PubMedQA, MedQA-USMLE, and MedMCQA under their original licenses.
提供机构：
Zenodo创建时间：
2026-07-07
