Uncertainty in Language Models: Assessment through Rank-Calibration

Xinmeng Huang, Shuo Li, Mengxin Yu +5

Language Models (LMs) have shown promising performance in natural language generation. However, as LMs often generate incorrect or hallucinated responses, it is crucial to correctl…