Self-Consistency Improves Chain of Thought Reasoning in Language Models

作者： Xuezhi Wang, Jason Wei, Dale Schuurmans, Quoc Le, Ed Chi, Sharan Narang, Aakanksha Chowdhery, Denny Zhou (2022)

arXiv： 2203.11171

TLDR（中文）

自洽性（Self-Consistency）是对 CoT 的重要改进：不再贪心解码单条推理链， 而是采样多条不同的推理路径，然后取答案出现最多的那个（多数投票）。 这简单的技巧在多个推理基准上将准确率提升了 10-20 个百分点。

TLDR (English)

Self-Consistency is a key improvement to CoT: instead of greedy decoding a single reasoning chain, sample multiple diverse reasoning paths and take the most frequent answer (majority vote). This simple trick improves accuracy by 10-20 percentage points on multiple reasoning benchmarks.