Chain-of-Thought Prompting Elicits Reasoning in Large Language Models

Wei J., Wang X., Schuurmans D., Bosma M., Ichter B., Xia F., Chi E. H., Le Q. V. and Zhou D. Chain-of-thought prompting elicits reasoning in large language models. NIPS, 2022.

概

Chain of Thought (CoT).

CoT

• 标准的 Prompt 如上左图所示, 而 CoT 则如右图所示. 不同之处在于, CoT 会给出一个 step by step 推理的例子, 这有助于激发模型的潜能 (当然, 前提是模型确实有这方面的潜能, CoT 比标准的 prompt 效果好通常要求模型规模 > 10B).

• 这篇文章 here 给了一个理论解释, 即为什么涉及到推理的问题是非常依赖 CoT 来分解问题并逐步回答的.

本栏目推荐文章
• 【略读论文|大模型相关】Zero-Shot Relational Learning on Temporal Knowledge Graphs with Large Language Models
• 南阳师范大学 外国语学院 Foreign language school
• 大模型评测-微软亚洲研究院：A Survey on Evaluation of Large Language Models论文分享
• Language Models are Unsupervised Multitask Learners
• InstructGPT《InstructGPT: Training language models to follow instructions with human feedback》解读
• GPT-3《Language Models are Few-Shot Learners》解读
• GPT-2 《Language Models are Unsupervised Multitask Learners》解读
• GPT-1论文《Improving Language Understanding by Generative Pre-Training》解读
• 《OneLLM: One Framework to Align All Modalities with Language》论文学习
• 《Learning Transferable Visual Models From Natural Language Supervision》论文学习

Chain-of-Thought Prompting Reasoning Language Thoughtchain-of-thought prompting reasoning language chain-of-thought prompting language thought chain-of-thought knowledge-intensive chain-of-thought interleaving chain-of-thought模型 思维 原理 cot chain-of-thought shot zero-shot reasoning language towards models language programming prompting models synergizing reasoning language acting recommendation personalized prompting language