直觉
一致性(知识库)
常识推理
计算机科学
解码方法
边距(机器学习)
人工智能
集合(抽象数据类型)
理论计算机科学
自然语言处理
认知科学
心理学
机器学习
算法
程序设计语言
作者
Xuezhi Wang,Jason Lee,Dale Schuurmans,Quoc V. Le,Ed H.,Denny Zhou
出处
期刊:Cornell University - arXiv
日期:2022-01-01
被引量:342
标识
DOI:10.48550/arxiv.2203.11171
摘要
Chain-of-thought prompting combined with pre-trained large language models has achieved encouraging results on complex reasoning tasks. In this paper, we propose a new decoding strategy, self-consistency, to replace the naive greedy decoding used in chain-of-thought prompting. It first samples a diverse set of reasoning paths instead of only taking the greedy one, and then selects the most consistent answer by marginalizing out the sampled reasoning paths. Self-consistency leverages the intuition that a complex reasoning problem typically admits multiple different ways of thinking leading to its unique correct answer. Our extensive empirical evaluation shows that self-consistency boosts the performance of chain-of-thought prompting with a striking margin on a range of popular arithmetic and commonsense reasoning benchmarks, including GSM8K (+17.9%), SVAMP (+11.0%), AQuA (+12.2%), StrategyQA (+6.4%) and ARC-challenge (+3.9%).
科研通智能强力驱动
Strongly Powered by AbleSci AI