Language Models are Few-Shot Learners

Authors: Brown et al. Year: 2020 ArXiv/Link: https://arxiv.org/abs/2005.14165

Summary

Demonstrates few-shot in-context learning without gradient updates as a capability of large language models.

Key Concepts

• Few-shot learning
• In-context learning
• Prompt engineering
• Task demonstration
• No gradient updates

Impact

Fundamental capability enabling practical prompting

Category

Prompting Techniques

---