EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Copiloting the Copilots for Automated Program Repair

Yuxiang Wei, Chunqiu Steven Xia, Lingming Zhang

Communications of the ACM · 01.04.2026 · 69 (4) · 89-94

DOI: 10.1145/3788082 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Communications of the ACM, online erschienen am 31.03.2026. Verfasst von Yuxiang Wei und 2 weiteren Autor:innen. Die Arbeit nennt 24 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

During Automated Program Repair (APR), it can be challenging to synthesize correct patches for real-world systems in general-purpose programming languages. Recent large language models (LLMs) have been shown to be helpful “copilots” in assisting developers with various coding tasks, and have also been directly applied for patch synthesis. However, most LLMs treat programs as sequences of tokens, meaning that they are ignorant of the underlying semantics constraints of the target programming language. This results in plenty of statically invalid generated patches, impeding the practicality of the technique. Therefore, we propose Repilot , a general code generation framework to further copilot the AI “copilots” (i.e., LLMs) by synthesizing more valid patches during the repair process. Our key insight is that many LLMs produce outputs autoregressively (i.e., token by token), resembling human writing programs, which can be significantly boosted and guided through a Completion Engine. Repilot synergistically synthesizes a candidate patch through the interaction between an LLM and a Completion Engine, which 1) prunes away infeasible tokens suggested by the LLM and 2) proactively completes the token based on the suggestions provided by the Completion Engine. Our evaluation on a subset of the widely-used Defects4j 1.2 and 2.0 datasets shows that Repilot outperforms state-of-the-art techniques by fixing 27% and 47% more bugs, respectively. Moreover, Repilot produces more valid and correct patches than the base LLM with the same budget. While we focus on leveraging Repilot for APR in this work, the overall approach is also generalizable to other code generation tasks.

Zitieren

APA

Wei, Y., Xia, C. S., & Zhang, L. (2026). Copiloting the Copilots for Automated Program Repair. Communications of the ACM, 69(4), 89-94. https://doi.org/10.1145/3788082

Vancouver

Wei Y, Xia CS, Zhang L. Copiloting the Copilots for Automated Program Repair. Communications of the ACM. 2026;69(4):89-94. doi:10.1145/3788082

BibTeX
@article{Wei2026,
  author = {Wei, Yuxiang and Xia, Chunqiu Steven and Zhang, Lingming},
  title = {Copiloting the Copilots for Automated Program Repair},
  journal = {Communications of the ACM},
  year = {2026},
  volume = {69},
  number = {4},
  pages = {89--94},
  doi = {10.1145/3788082}
}
Kostenlos starten →