Large language model uncertainty proxies: discrimination and calibration for medical diagnosis and treatment.
The inability of large language models (LLMs) to communicate uncertainty is a significant barrier to their use in medicine. Before LLMs can be integrated into patient care, the field must assess methods to estimate uncertainty in ways that are useful to physician-users.
Author(s): Savage, Thomas, Wang, John, Gallo, Robert, Boukil, Abdessalem, Patel, Vishwesh, Safavi-Naini, Seyed Amir Ahmad, Soroush, Ali, Chen, Jonathan H
DOI: 10.1093/jamia/ocae254