1 paper · 1 filter
Yue Xin, Chen Shen, Shaotian Yan +5
Chain-of-Thought (CoT) prompting enhances the math reasoning capability of large language models (LLMs) to a large margin. However, the mechanism underlying such improvements remai…