Evaluating the Generalizability of LLMs in Automated Program Repair
This work addresses the limited generalizability of LLMs in program repair for software engineering, highlighting an incremental evaluation gap.
The study evaluated the generalizability of 11 top-performing LLMs in automated program repair, finding that their performance decreased by 49.48% and 42.90% on a new dataset, and prompt engineering improved results but was insufficient to match original performance.
LLM-based automated program repair methods have attracted significant attention for their state-of-the-art performance. However, they were primarily evaluated on a few well known datasets like Defects4J, raising questions about their effectiveness on new datasets. In this study, we evaluate 11 top-performing LLMs on DEFECTS4J-TRANS, a new dataset derived from transforming Defects4J while maintaining the original semantics. Results from experiments on both Defects4J and DEFECTS4J-TRANS show that all studied LLMs have limited generalizability in APR tasks, with the average number of correct and plausible patches decreasing by 49.48% and 42.90%, respectively, on DEFECTS4J-TRANS. Further investigation into incorporating additional repair-relevant information in repair prompts reveals that, although this information significantly enhances the LLMs' capabilities (increasing the number of correct and plausible patches by up to 136.67% and 121.82%, respectively), performance still falls short of their original results. This indicates that prompt engineering alone is insufficient to substantially enhance LLMs' repair capabilities. Based on our study, we also offer several recommendations for future research.