SWT-Bench Verified

Papers using SWT-Bench Verified (39)

Frontier Coding Agents Use Metaprogramming to Adapt to Unfamiliar Programming Languages2026Function-Aware Fill-in-the-Middle as Mid-Training for Coding Agent Foundation Models2026Socratic-SWE: Self-Evolving Coding Agents via Trace-Derived Agent Skills2026BoostAPR: Boosting Automated Program Repair via Execution-Grounded Reinforcement Learning with Dual Reward Models2026Yet Even Less Is Even Better For Agentic, Reasoning, and Coding LLMs2026From SWE-ZERO to SWE-HERO: Execution-free to Execution-based Fine-tuning for Software Engineering Agents2026SWE-Fuse: Empowering Software Agents via Issue-free Trajectory Learning and Entropy-aware RLVR Training2026Your Code Agent Can Grow Alongside You with Structured Memory2026CodeScout: An Effective Recipe for Reinforcement Learning of Code Search Agents2026KAT-Coder-V2 Technical Report2026SWE-Universe: Scale Real-World Verifiable Environments to Millions2026SWE-Master: Unleashing the Potential of Software Engineering Agents via Post-Training2026SWE-World: Building Software Engineering Agents in Docker-Free Environments2026Pull Requests as a Training Signal for Repo-Level Code Editing2026SWE Context Bench: A Benchmark for Context Learning in Coding2026Immersion in the GitHub Universe: Scaling Coding Agents to Mastery2026Hybrid-Gym: Training Coding Agents to Generalize Across Tasks2026SWE-Prot\'eg\'e: Learning to Selectively Collaborate With an Expert Unlocks Small Language Models as Software Engineering Agents2026SWE-Tester: Training Open-Source LLMs for Issue Reproduction in Real-World Repositories2026Learning Adaptive Parallel Execution for Efficient Code Localization2026SWE-Replay: Efficient Test-Time Scaling for Software Engineering Agents2026SWE-EVO: Benchmarking Coding Agents in Long-Horizon Software Evolution Scenarios2025Self-Abstraction from Grounded Experience for Plan-Guided Policy Refinement2025InfCode: Adversarial Iterative Refinement of Tests and Patches for Reliable Software Issue Resolution2025CWM: An Open-Weights LLM for Research on Code Generation with World Models2025REFINE: Enhancing Program Repair Agents through Context-Aware Patch Refinement2025Saving SWE-Bench: A Benchmark Mutation Approach for Realistic Agent Evaluation2025BugPilot: Complex Bug Generation for Efficient Learning of SWE Skills2025Reinforcement Learning-Guided Chain-of-Draft for Token-Efficient Code Generation2025CoreCodeBench: Decoupling Code Intelligence via Fine-Grained Repository-Level Tasks2025AssertFlip: Reproducing Bugs via Inversion of LLM-Generated Passing Tests2025SWE-Dev: Building Software Engineering Agents with Training and Inference Scaling2025Training Long-Context, Multi-Turn Software Engineering Agents with Reinforcement Learning2025Guided Search Strategies in Non-Serializable Environments with Applications to Software Engineering Agents2025SWE-rebench: An Automated Pipeline for Task Collection and Decontaminated Evaluation of Software Engineering Agents2025Satori-SWE: Evolutionary Test-Time Scaling for Sample-Efficient Software Engineering2025FeedbackEval: A Benchmark for Evaluating Large Language Models in Feedback-Driven Code Repair Tasks2025SEAlign: Alignment Training for Software Engineering Agent2025Thinking Longer, Not Larger: Enhancing Software Engineering Agents via Scaling Test-Time Compute2025