Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting

Devs

Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting | Read Paper on Bytez