arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.04449cs.SE
收藏

基准中有什么?SWE-Bench在自动程序修复中的案例

What's in a Benchmark? The Case of SWE-Bench in Automated Program Repair

Universitat Politècnica de Catalunya(加泰罗尼亚理工大学)

Matias Martinez, Xavier Franch

AI总结:

本文研究了SWE-Bench排行榜上的提交来源、LLM使用情况及行业参与度,揭示了专有LLM在自动程序修复中的主导地位。

摘要:

自动程序修复(APR)的快速发展得益于人工智能领域的进步,特别是大语言模型(LLMs)和基于代理的系统。SWE-Bench是一个旨在通过从流行开源Python仓库中挖掘的真实问题来评估修复系统的方法。其公开的排行榜——SWE-Bench Lite和Verified——已成为跟踪进展和比较解决方案的核心平台。在本文中,我们对这两个排行榜进行了首次全面研究,分析了提交解决方案的人群、提交物、使用的LLMs以及方法的开放性。我们分析了提交给Lite排行榜的79项和提交给Verified排行榜的133项。我们的结果显示,这两个排行榜上的大多数提交来自行业,特别是小型公司和大型上市公司。这些提交往往取得顶尖成绩,尽管学术贡献——通常开源——仍然具有竞争力。我们还发现专有LLMs在占主导地位,特别是Claude家族,目前在两个排行榜上取得最先进结果的是Claude 4 Sonnet。这些发现为SWE-Bench生态系统提供了见解,可以指导未来基于基准的研究中的更大透明度和多样性。

英文摘要:

The rapid progress in Automated Program Repair (APR) has been fueled by advances in AI, particularly large language models (LLMs) and agent-based systems. SWE-Bench is a benchmark designed to evaluate repair systems using real issues mined from popular open-source Python repositories. Its public leaderboards-SWE-Bench Lite and Verified-have become central platforms for tracking progress and comparing solutions. In this paper, we present the first comprehensive study of these two leaderboards, examining who is submitting solutions, the products behind the submissions, the LLMs employed, and the openness of the approaches. We analyze 79 entries submitted to Lite leaderboard and 133 to Verified. Our results show that most entries on both leaderboards originate from industry, particularly small companies and large publicly traded companies. These submissions often achieve top results, although academic contributions-typically open source-also remain competitive. We also find a clear dominance of proprietary LLMs, especially Claude family, with state-of-the-art results on both leaderboards currently achieved by Claude 4 Sonnet. These findings offer insights into the SWE-Bench ecosystem that can guide greater transparency and diversity in future benchmark-driven research.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑