Anran Li, Yuanyuan Chen, Wenjun Long +16
Large language models (LLMs) are increasingly adapted for medical applications, but most are trained using data from a single institution because privacy and governance constraints…