Too Long; Didn’t Read

This paper explores the application of large language models (LLMs) in genetic improvement (GI) within software engineering through experimental evaluation. Insights reveal the impact of LLMs on search-based techniques, program repair, and code generation, with a focus on the JCodec project. Discover how LLM mutations enhance software development processes.