Bugs in Large Language Models Generated Code: An Empirical Study Revealing Distinctive Bug Patterns

LLM-generated code exhibits various bug patterns, including Misinterpretation, Missing Corner Cases, and Non-Prompted Consideration, highlighting potential issues with automatic code generation.