In-Context Learning

Quick Definition

LLMs learning new tasks from prompt examples without weight updates

The ability of LLMs to learn new tasks from examples provided in the prompt without updating model weights.

few-shot classification, task adaptation

few-shot-learning large-language-model

Detecting human body positions and joint locations

Partitioning images into labeled segments per pixel

Combining LLMs with external knowledge for accurate responses

Converting written text into natural-sounding spoken audio

Creating smaller models that replicate larger model behavior

AI field enabling computers to understand human language