In-Context Learning

Large Language Models Are Implicitly Topic Models: Explaining and Finding Good Demonstrations for In-Context Learning

The Thirty-seventh Conference on Neural Information Processing Systems (NeurIPS 2023)