What is few-shot prompting?
Also called: few-shot learning, examples in prompts
Few-shot prompting means including a few examples of the input and the output you want inside your prompt, so the AI copies the pattern. It's one of the simplest ways to get consistent formats, tone or labels without any training.
How does few-shot prompting work?
You show two to five worked examples, then give the new input. The model picks up the pattern from the examples and applies it. 'Zero-shot' means asking with no examples at all.
Why does few-shot prompting matter?
Examples often work better than long descriptions. They're especially useful for classification, formatting and matching a house style.
An example
Example: Show the AI three customer reviews labelled positive, negative or mixed, then paste a new review, and it labels it the same way.
Related terms
Learn it properly
Few-shot prompting is covered in Use AI Well, with short lessons, practice questions and instant feedback.