---
title: 'Chain-of-thought prompting'
description: 'Prompting a model with worked intermediate reasoning, or asking it to generate intermediate steps, before producing an answer.'
canonical_url: 'https://darkfactory.dev/glossary/chain-of-thought-prompting'
markdown_url: 'https://darkfactory.dev/glossary/chain-of-thought-prompting.md'
collection: glossary
date_published: '2026-08-03T00:00:00-04:00'
date_modified: '2026-08-26T00:00:00-04:00'
---

# Chain-of-thought prompting


## Definition

Chain-of-thought prompting supplies worked examples with intermediate reasoning steps, asks the model to produce such steps, or combines both approaches. The generated steps can help a model decompose some arithmetic, symbolic, and commonsense problems before it commits to an answer.

Jason Wei and coauthors gave the method its current name in their January 2022 paper, *Chain-of-Thought Prompting Elicits Reasoning in Large Language Models*. Their experiments used few-shot examples containing a question, a chain of intermediate steps, and a final answer. The reported gains were empirical results for the models and tasks tested, not proof that visible reasoning always improves an answer.

## What the reasoning text proves

A chain of thought is model output, not a trustworthy execution log. Miles Turpin and coauthors showed that models can produce plausible explanations for answers influenced by hidden biasing features while failing to mention that influence. The reasoning may still be useful as a draft, a decomposition, or an object to critique, but it requires the same independent checks as the final answer.

Applications also differ in whether they expose reasoning text, keep a private scratchpad, or return only a concise answer. Those interface choices should not be mistaken for evidence about how the model internally computed the result.

## Distinguish it from nearby terms

- Few-shot prompting provides examples. It becomes chain-of-thought prompting when those examples include intermediate reasoning.
- A rationale explains or justifies an answer. It may be generated after the answer and need not have caused it.
- Verification tests the answer against evidence, computation, or another independent check. A fluent chain of thought is not verification.

## Check your understanding

A model gives the correct numerical answer with a persuasive derivation. Identify one check that tests the calculation without relying on the model's own explanation.

## Related terms

- [Reasoning model](https://darkfactory.dev/glossary/reasoning-model)
- [Prompt chaining](https://darkfactory.dev/glossary/prompt-chaining)

## Evidence and further reading

- [Google Machine Learning Glossary](https://developers.google.com/machine-learning/glossary/)
- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://arxiv.org/abs/2201.11903)
- [Language Models Don't Always Say What They Think](https://arxiv.org/abs/2305.04388)