1 paper · 1 filter
Donald Ye, Max Loffgren, Om Kotadia +2
Chain-of-Thought (CoT) explanations are widely used to interpret how language models solve complex problems, yet it remains unclear whether these step-by-step explanations reflect…