Language models are few-shot learners

In three linesOpenAI publishes foundational research on few-shot learning capabilities in language models. LLMs can perform tasks with minimal examples without fine-tuning, revealing emergent rapid adaptation capacity.

Summary generated by Claude — human-verified