#01
CliniCARE-Bench: Clinical Calibrated Audit of Medical Reasoning in EHR
Large language models perform strongly on medical knowledge benchmarks, but reliable clinical deployment requires agents to conduct defensible investigations over heterogeneous, longitudinal records: determining what evi
推荐理由CliniCARE-Bench: Clinical Calibrated：Large language models perform strongly on medical knowledge benchmarks, but reliable。