ReCatcher systematically measures regressions in LLM code generation across correctness, static quality, and performance, and its evaluation shows fine-tuning, merging, and new releases each introduce specific regressions.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SE 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
ReCatcher: Towards LLMs Regression Testing for Code Generation
ReCatcher systematically measures regressions in LLM code generation across correctness, static quality, and performance, and its evaluation shows fine-tuning, merging, and new releases each introduce specific regressions.