SWE-Edit: Rethinking Code Editing for Efficient SWE-Agent
Microsoft(微软);Department of Computer Sciences, University of Wisconsin–Madison(威斯康星大学麦迪逊分校计算机科学系);Stanford Institute for Human-Centered Artificial Intelligence (HAI), Stanford University(斯坦福大学人机交互研究所)
Large language model agents have made strong progress on software engineering, yet current systems suffer from a context coupling problem: the standard code editing interface conflates code inspection, modification planning, and edit execution within a single context window, forcing agents to interleave exploratory viewing with strictly formatted edit generation. Irrelevant context accumulates and edit reliability degrades. We propose SWE-Edit, which decomposes the editing interface into two specialized subagents: a Viewer that extracts task-relevant code on demand, and an Editor that executes modifications from high-level natural language plans -- letting the main agent focus on reasoning while delegating context-intensive operations to clean context windows. On SWE-Bench Verified, this decomposition raises resolve rate by 2.1 pp and cuts inference cost by 17.9%, with consistent gains across multiple reasoning-model families (Kimi-K2, MiniMax-M2.1, GLM-4.7). We further show that effective edit-format selection can be trained into a small model rather than requiring frontier-scale capacity: GRPO training on Qwen3-8B with an adaptive find-replace/whole-file-rewrite policy improves edit success by 12.5 pp and brings an 8B open-source editor to parity with GPT-5-nano on downstream SWE-Bench resolve rate. To enable rapid editor iteration, we release PR-Edit, a lightweight evaluation whose scores correlate strongly with SWE-Bench resolve rate. We release our code at https://github.com/microsoft/SWE-Edit.
Related Knowledge Perturbation Matters: Rethinking Multiple Pieces of Knowledge Editing in Same-Subject
Institute of Computing Technology, Chinese Academy of Sciences(中国科学院计算技术研究所);University of Chinese Academy of Sciences(中国科学院大学);People’s Public Security University of China(中国人民公安大学)
The Labyrinth and the Thread: Rethinking Regularizations in Sequential Knowledge Editing for Large Language Models
Bosch Center for Artificial Intelligence (BCAI)(博世人工智能中心(BCAI));Bosch (China) Investment Ltd.(博世(中国)投资有限公司);School of Statistics, East China Normal University(东华大学统计学院)
Rethinking Machine Unlearning for Large Language Models
Michigan State University(密歇根州立大学);IBM Research(IBM研究院);University of North Carolina at Chapel Hill(北卡罗来纳大学教堂山分校);University of California, Santa Cruz(加州大学圣克鲁兹分校);ByteDance Research(字节跳动研究院);Stanford University(斯坦福大学)
SHERLOC: Structured Diagnostic Localization for Code Repair Agents
NVIDIA(英伟达);Santa Clara, CA 95051, USA(美国加利福尼亚州圣克拉拉);TU Darmstadt(达姆施塔特工业大学);National Research Center for Applied Cybersecurity ATHENE(国家应用网络安全研究中心 ATHENE)
Pull Requests as a Training Signal for Repo-Level Code Editing
King's College London, UK(伦敦国王学院);Chinese University of Hong Kong, HK(香港中文大学);National University of Singapore, SG(新加坡国立大学);Microsoft Research Asia, CN(微软亚洲研究院);The Alan Turing Institute, UK(艾伦·图灵研究所)
BugPilot: Complex Bug Generation for Efficient Learning of SWE Skills
Cornell University(康奈尔大学);University of California San Diego(加州大学圣地亚哥分校);University of North Carolina at Chapel Hill(北卡罗来纳大学教堂山分校);Microsoft Research(微软研究院)
A Single Patch Is Not Enough: Deterministic Fusion of Repair Candidates
Yanshan University(燕山大学);Beijing Guoyan Network Data Technology Co., Ltd.(北京国燕网络数据技术有限公司);Peking University(北京大学);Google(谷歌);University of Melbourne(墨尔本大学);University of Luxembourg(卢森堡大学);Aalto University(阿尔托大学)
Fail-Fast, Restart-Smart: Early Failure Prediction and Restart for SWE Agentic Tasks
Singapore Management University(新加坡管理大学);University of Alberta(阿尔伯塔大学);Nanjing University of Science and Technology(南京理工大学);Tel Aviv University(特拉维夫大学)