Teaching Large Language Models to Self-Debug

Xinyun Chen, Maxwell Lin, Nathanael Schärli +1

Large language models (LLMs) have achieved impressive performance on code generation. However, for complex programming tasks, generating the correct solution in one go becomes chal…