AI Glossary

Few-shot prompting

Also called few-shot, in-context examples

What does “Few-shot prompting” mean?

Few-shot prompting: Including a handful of example inputs with their correct outputs in the prompt so the model can copy the pattern. The model is not retrained; it picks up the format, style or labeling rules from the examples it sees in that request.

Few-shot prompting visual guide: Including a handful of example inputs with their correct outputs in the prompt so the model can copy the pattern. The model is not retrained; it picks up the format, style or labeling rules from the examples it sees in that request.
Visual guide Open full size (new tab)

What is an example of “Few-shot prompting”?

Showing three sample support tickets already tagged "bug," "billing" or "feature request" helps the model tag a fourth one the same way.

Keep learning

See also

Zero-shot promptingTermIn-context learningTermPrompt engineeringTerm

More from Franklin

On Franklin

Prompt libraryPage

Same topic

Frequency and presence penaltiesTermEmbeddingsTermGroundingTermContext windowTermHallucinationTerm