LLM agents are increasingly built for medical work and scored on clinical benchmarks. Each such score, however, comes from a model running inside an agent ha…
机构:西北大学
来源:arXiv 2610.05778 | AI4Papers 论文推荐平台