tanhaosheng/

surgeon-tested-clinical-ai-benchmark

An independent, reproducible evaluation of large language models (LLMs) on real, de-identified cancer cases — scored by a practicing surgeon item-by-item against current clinical guidelines.

Homepage & full leaderboard: https://tanhaosheng.asia Methodology (citable authority, TH-CAB v1.1): https://tanhaosheng.asia/methodology/ Open data layer: https://tanhaosheng.asia/data/

This is a benchmark / research dataset, not clinical… See the full description on the dataset page: https://huggingface.co/datasets/tanhaosheng/surgeon-tested-clinical-ai-benchmark.