遇见数据集

RxSafeBench

收藏
DataCite Commons2025-06-17 更新2025-09-08 收录
官方服务:

资源简介:

RxSafeBench.zip is a benchmark dataset designed to evaluate the medication safety capabilities of large language models (LLMs) in simulated clinical consultation scenarios. The final benchmark consists of 2,443 high-quality consultation scenarios, evenly split across different types of medication risks. This dataset is essential for advancing research on LLM-based systems in healthcare, particularly in improving medication safety.

提供机构:
figshare
创建时间:
2025-06-17
二维码
社区交流群
二维码
科研交流群
商业服务