arXiv · ML· Amirali Sajadi, Kostadin Damevski, Preetha Chatterjee·· 8 小时前AI 评分52
How Safe Are AI-Generated Patches? A Large-scale Study on Security Risks in LLM and Agentic Automated Program Repair on SWE-bench
How Safe Are AI-Generated Patches? A Large-scale Study on Security Risks in LLM and Agentic Automated Program Repair on SWE-bench
AI 导读
研究人员对基于SWE-bench的LLM和Agent自动程序修复（APR）进行了大规模安全分析，使用20000+个GitHub issue对比了开发者补丁与AI生成补丁的安全风险。
来源：arXiv · ML · arxiv.org