SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering

Authors: John Yang, Carlos E. Jimenez, Alexander Wettig, Kilian Lieret, Shunyu Yao, Karthik Narasimhan, Ofir Press (2024)

arXiv: 2405.15793

Domains

Applications

TLDR (English)

Proposes ACI (Agent-Computer Interface) concept, emphasizing "what tools/interface agent uses ≥ what model used". GPT-4 + good ACI improves SWE-bench 6x, establishing coding agent engineering methodology.

TLDR（中文）

提出 ACI（Agent-Computer Interface）概念，强调"agent 用什么工具/界面 ≥ 用什么模型"。把 GPT-4 + 良好 ACI 在 SWE-bench 提升 6 倍，奠定 coding agent 工程方法论。

Appears in These Articles

• Agent 与工具使用：模型不只是聊天
• 代码生成：模型如何写程序
• Agents and Tool Use: Models Are More Than Chat
• Code Generation: How Models Write Programs

Co-cited Papers

These papers appear in the same articles as this one

• Model Context Protocol (MCP) Anthropic (2024)
• Evaluating Large Language Models Trained on Code Mark Chen et al. (2021)
• DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning DeepSeek-AI (2025)
• SWE-bench: Can Language Models Resolve Real-World GitHub Issues? Carlos E. Jimenez et al. (2024)

Related Papers

Other papers in the same domain

• Model Context Protocol (MCP) Anthropic (2024)
  Applications
• SWE-bench: Can Language Models Resolve Real-World GitHub Issues? Carlos E. Jimenez et al. (2024)
  EvaluationApplications
• Toolformer: Language Models Can Teach Themselves to Use Tools Timo Schick et al. (2023)
  Applications
• Reflexion: Language Agents with Verbal Reinforcement Learning Noah Shinn et al. (2023)
  Applications
• Efficient Guided Generation for Large Language Models Brandon T. Willard et al. (2023)
  InferenceApplications
• Tree of Thoughts: Deliberate Problem Solving with Large Language Models Shunyu Yao et al. (2023)
  ReasoningApplications