AI Blog

AI Blog

by Michele Laurelli

Zero-Shot Learning

/ˈzɪroʊ ʃɒt ˈlɜːrnɪŋ/
Concept
Definition

A model's ability to perform tasks it wasn't explicitly trained on, using only task descriptions or examples.

Zero-shot learning leverages pre-trained knowledge to handle new tasks without additional training. Large language models excel at this through prompt engineering.

Examples

1

GPT-3 translating to languages not in training

2

CLIP matching images to unseen concepts

3

ChatGPT solving novel reasoning tasks

Michele Laurelli - AI Research & Engineering