SaySelf: A machine learning training framework that teaches LLMs to express more accurate and detailed confidence estimates
Language learning models (LLMs), who are very good at reasoning and giving good answers, are sometimes honest about their mistakes ...
