Dependency upgrades can break downstream projects without changing the library interface. Such behavioral breaking changes are difficult for developers to fix, because the failing test does not always point to the root API, the upstream API that causes the break. Existing LLM-based repair methods obtain evidence for th...
Jun-Tao Chen, Ting-Ting Bi, Yan-Lin Wang et al.· 0 citations
This work presents XAgent, an execution-guided agentic framework that analyzes dynamic behavior and additional program context to localize and validate issues, opening new opportunities for LLM-based coding agents to achieve more robust and generalizable software maintenance.
H. Huynh, Patanamon Thongtanunam, Michael Fu et al.· 0 citations
This study proposes local Platt-scaling applied separately to three different fine-grained confidence scores, finding that fine-grained confidence scores consistently achieve lower calibration error across a broader range of probability intervals, and this effect is further amplified when global Platt-scaling is applie...
The ablation study confirms that the three context facets are mutually complementary, and that multi-agent scaffolding and base-model capacity each play an essential role for agentic vulnerability repair.
Michael Fu, Qiyue Mei, Patanamon Thongtanunam et al.· arXiv.org· 1 citation
An empirical study of agentic code reviews using CodeRabbit as a case study shows that agentic reviews receive mixed reception, and finds that lightweight learning-based methods achieve up to 76% F1 score, suggesting learnable patterns exist between code reviews and their corresponding feedback.
