Chain of Thought
A prompting technique that instructs the model to reason step-by-step before answering.
Chain-of-thought (CoT) prompting adds 'Let's think step by step' or provides examples of detailed reasoning before a final answer. This dramatically improves performance on multi-step math, logic, and reasoning problems. Extended thinking models (like Claude 3.7) apply CoT internally before producing a response, trading latency and token cost for higher accuracy.
Related Terms
The practice of designing inputs to elicit optimal outputs from language models.
Providing a small number of input-output examples in the prompt to guide the model.
Prompting a model to perform a task without providing any examples.
A model variant that produces explicit step-by-step thinking before answering.