Large Language Models are Zero-Shot Reasoners

Authors: Takeshi Kojima, Shixiang Shane Gu, Machel Reid, Yutaka Matsuo, Yusuke Iwasawa (2022)

arXiv: 2205.11916

Domains

Reasoning

TLDR (English)

A single phrase "Let's think step by step" boosts math accuracy from ~17% to ~78%. CoT capability is inherent in models, triggered by prompts—this discovery shocked the entire community.

TLDR（中文）

一句 "Let's think step by step" 就能在数学题上把准确率从 ~17% 拉到 ~78%。CoT 能力是模型自带、prompt 触发，这一发现震惊了整个社区。

Appears in These Articles

• 提示工程：与模型对话的艺术
• Prompt Engineering: The Art of Talking to Models

Co-cited Papers

These papers appear in the same articles as this one

• DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning DeepSeek-AI (2025)
• Let's Verify Step by Step Hunter Lightman et al. (2023)
• Self-Consistency Improves Chain of Thought Reasoning in Language Models Xuezhi Wang et al. (2022)
• Chain-of-Thought Prompting Elicits Reasoning in Large Language Models Jason Wei et al. (2022)

Related Papers

Other papers in the same domain

• DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning DeepSeek-AI (2025)
  ReasoningAlignment
• OpenAI o1 System Card OpenAI (2024)
  ReasoningSafety
• Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters Charlie Snell et al. (2024)
  ReasoningInference
• Let's Verify Step by Step Hunter Lightman et al. (2023)
  ReasoningEvaluation
• Tree of Thoughts: Deliberate Problem Solving with Large Language Models Shunyu Yao et al. (2023)
  ReasoningApplications
• Self-Consistency Improves Chain of Thought Reasoning in Language Models Xuezhi Wang et al. (2022)
  Reasoning