Home/Glossary/Zero-Shot / Few-Shot Learning

AI Glossary

Zero-Shot / Few-Shot Learning

Teaching AI by example — or without any examples

Definition

Zero-shot learning is asking a model to perform a task it was not explicitly trained on, without any examples. Few-shot learning provides a handful of input-output examples in the prompt to guide behavior. These capabilities emerge from large-scale pre-training and are key reasons modern LLMs are so versatile.

Related Terms

Back to Glossary