1 paper · 1 filter
James Chua, Edward Rees, Hunar Batra +4
Chain-of-thought prompting (CoT) has the potential to improve the explainability of language model reasoning. But CoT can also systematically misrepresent the factors influencing m…