What is chain-of-thought prompting?
Getting the model to write out intermediate reasoning steps before its answer.
- Having the model write intermediate reasoning steps ✔
- Chaining several models in a row
- Linking prompts with hyperlinks
Why: A 2022 Google paper showed that example answers with worked steps make large models much better at math and logic. Writing the steps gives the model more tokens to compute with before committing to an answer.
Huh, didn't know that: With just eight chain-of-thought examples, a 540 billion parameter model reached state of the art accuracy (in 2022) on the GSM8K math word problems.
My math teacher invented chain-of-thought prompting decades ago. She called it show your work.
References
Answer: Having the model write intermediate reasoning steps. A 2022 Google paper showed that example answers with worked steps make large models much better at math and logic. Writing the steps gives the model more tokens to compute with before committing to an answer.
- Wei et al., Chain-of-Thought Prompting Elicits Reasoning in Large Language Models (arXiv 2201.11903) arxiv.org/abs/2201.11903
a series of intermediate reasoning steps -- significantly improves the ability of large language models to perform complex reasoning
Checked 2026-10-10.
Think this answer is wrong? Tap "Challenge this answer" on the card in the game and tell us why.
More stuff everyone should know
That's stuff everyone should know.