CLEAR: Cross-Source Evidence Adjudication for Large Language Models in Medicine

Shuai Wang, Yize Zhao, Qingyu Chen

Medical knowledge evolves continuously, whereas the parametric knowledge encoded in large language models (LLMs) is fixed at training time. External retrieval, including retrieval-…