用双模型策略减少自动化修代码的无效提案，提升工业级部署可信度。
Abstain and Validate: A Dual-LLM Policy for Reducing Noise in Agentic Program Repair
- 引入弃权与验证双策略，提前过滤难修复的漏洞和差质量补丁。
- 在174个真实漏洞上，综合使用可使成功率提升最多39个百分点。
- 适合需要高可靠性、减少人工返工的工业级自动化代码修复场景。
自动化程序修复（APR）正日益用于解决工业界中的复杂、仓库级漏洞，但最终修复补丁仍需人工审查以确保有效。若频繁提交低效或无效补丁，将造成大量噪声，浪费开发者时间并削弱对自动修复的信任。本文提出两种互补的基于大语言模型（LLM）的策略：虫洞弃权（bug abstention）用于排除系统难以修复的漏洞，补丁验证（patch validation）用于拒绝不符合要求的修复方案。我们在谷歌代码库的三组漏洞上评估了这两种策略，其候选补丁由内部的智能体式APR系统生成。针对174个由人工报告的漏洞，采用策略后可分别提升成功率最高达13个百分点（弃权）和15个百分点（验证），两者结合时最高提升达39个百分点。对于由机器生成报告的空指针异常和沙箱检测出的漏洞，补丁验证也提升了平均单样本成功率达到显著水平。该双策略为实现可靠、可扩展的工业级智能体式自动修复系统提供了实用路径。
原文摘要 · Abstract (English)
Agentic Automated Program Repair (APR) is increasingly tackling complex, repository-level bugs in industry, but ultimately these patches still need to be reviewed by a human before committing them to ensure they address the bug. Showing patches unlikely to be accepted can lead to substantial noise, wasting valuable developer time and eroding trust in automated code changes. We introduce two complementary LLM-based policies to reduce such noise: bug abstention and patch validation policies. Bug abstention excludes bugs that the agentic APR system is unlikely to fix. Patch validation rejects patches that are unlikely to be a good fix for the given bug. We evaluate both policies on three sets of bugs from Google's codebase, and their candidate patches generated by an internal agentic APR system. On a set of 174 human-reported bugs, removing bugs and patches rejected by our policies can raise success rates by up to 13 percentage points and 15 percentage points, respectively, and by up to 39 percentage points in combination. On null pointer exceptions and sanitizer-reported bugs with machine-generated bug reports, patch validation also improves average single-sample success rates. This two-policy approach provides a practical path to the reliable, industrial-scale deployment of agentic APR systems.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。