Language models are few-shot learners

“Language models are few-shot learners” is a landmark research work presented at NeurIPS 2020, introducing and investigating the remarkable capability of large-scale language models to generalize to new tasks with very little supervision. The paper demonstrates that language models, such as GPT-3, can efficiently learn and perform a wide range of natural language tasks with only a handful of examples, fundamentally advancing state-of-the-art in general AI and NLP research. It is aimed at AI researchers, developers, and anyone interested in the frontier of artificial intelligence and language model capabilities.