Chain-of-thought
Also known as: CoT · think step by step
Chain-of-thought (CoT) is prompting the model to lay out its reasoning before giving the final answer. It’s a large, well-documented boost on multi-step problems.
- “Let’s think step by step.”
- “Show your work, then give the answer.”
- Few-shot CoT: include examples that already show reasoning.
It works because breaking a problem into steps lets the model use intermediate results.
The single best reliability trick: for anything multi-step, asking for the reasoning first is the highest-leverage prompt change you can make.