Cross-Examiner: Evaluating Consistency of Large Language Model-Generated Explanations

Danielle Villa, Maria Chang, Keerthiram Murugesan +2

Large Language Models (LLMs) are often asked to explain their outputs to enhance accuracy and transparency. However, evidence suggests that these explanations can misrepresent the…