Refining bug reports with hierarchical evidence from an unverified agent trajectory plus repository review raises Pass@1 on SWE-Bench Lite from 41% to 59.67% (GPT-5-mini) and similarly for other agents.
EXPEREPAIR: Dual-Memory Enhanced LLM-based Repository-Level Program Repair
9 Pith papers cite this work. Polarity classification is still indexing.
9
Pith papers citing it
abstract
Automatically repairing software issues remains a fundamental challenge at the intersection of software engineering and AI. Although recent advances in Large Language Models (LLMs) have demonstrated potential for repository-level repair tasks, current methods exhibit two notable limitations: (1) they often address issues in isolation, neglecting to incorporate insights from previously resolved issues, and (2) they rely on static, rigid prompting strategies that constrain their ability to generalize across diverse and evolving contexts. We propose ExpeRepair, a novel LLM-based program repair framework inspired by the dual-memory systems of human cognition, where episodic and semantic memory synergistically support learning and decision-making. Unlike existing methods, ExpeRepair continuously learns from historical repair experiences via dual-channel knowledge accumulation, enabling it to adaptively reuse past knowledge during inference. Specifically, ExpeRepair organizes prior repair knowledge into two complementary memories: an episodic memory that stores concrete repair demonstrations, and a semantic memory that encodes abstract, reflective insights. At inference time, ExpeRepair activates both memory systems by retrieving relevant demonstrations from episodic memory and recalling high-level repair insights from semantic memory. It further enhances adaptability through dynamic prompt composition, integrating both memory types to replace static prompts with context-aware, experience-driven prompts. We evaluate ExpeRepair on two benchmarks: SWE-Bench Lite and SWE-Bench Verified. Experimental results show that ExpeRepair achieves pass@1 scores of 60.3% and 74.6% on the two benchmarks, respectively, achieving the best performance among the evaluated open-source methods. We have open-sourced ExpeRepair at https://github.com/ExpeRepair/ExpeRepair.
citation-role summary
background 1
citation-polarity summary
years
2026 9roles
background 1polarities
background 1representative citing papers
Loc2Repair framework evaluation finds that file-level localization boosts LLM repo repair resolved rates by up to 7.7 percentage points on SWE-bench Verified.
Adding visual dependency-graph images to a text-based coding agent cuts token consumption by up to 26% while keeping issue-resolution accuracy roughly unchanged.
MocklessTester generates mockless Java unit tests using context mining and multi-level constraint fixing, improving coverage and mutation scores over baselines on Defects4J and Deps4J.
Agent-CoEvo is a multi-agent LLM framework that coevolves code patches and test patches to resolve repository-level issues, outperforming fixed-test baselines on SWE-bench Lite and SWT-bench Lite.
Embedding dynamic analysis into an LLM repair agent yields a claimed 79.4% resolution rate on SWE-bench Verified while cutting token use by about 25%.
A multi-granularity program graph with intra-procedural def-use edges and a data-flow slicing tool raises SWE-agent resolve rate on SWE-bench Lite by 4.7 points under a fixed open-source backbone.
Agent-generated tests mainly act as observational feedback channels and do not meaningfully improve issue resolution success in current LLM software engineering agents.
citing papers explorer
-
Bug Report Specification Refinement with Trajectory Guidance for Automated Program Repair
Refining bug reports with hierarchical evidence from an unverified agent trajectory plus repository review raises Pass@1 on SWE-Bench Lite from 41% to 59.67% (GPT-5-mini) and similarly for other agents.
-
Loc2Repair: A Framework for Evaluating the Impact of File-Level Issue Localization in Repo-Level LLM Repair
Loc2Repair framework evaluation finds that file-level localization boosts LLM repo repair resolved rates by up to 7.7 percentage points on SWE-bench Verified.
-
LLM Agents Can See Code Repositories
Adding visual dependency-graph images to a text-based coding agent cuts token consumption by up to 26% while keeping issue-resolution accuracy roughly unchanged.
-
LLM-based Mockless Unit Test Generation for Java
MocklessTester generates mockless Java unit tests using context mining and multi-level constraint fixing, improving coverage and mutation scores over baselines on Defects4J and Deps4J.
-
Beyond Fixed Tests: Repository-Level Issue Resolution as Coevolution of Code and Behavioral Constraints
Agent-CoEvo is a multi-agent LLM framework that coevolves code patches and test patches to resolve repository-level issues, outperforming fixed-test baselines on SWE-bench Lite and SWT-bench Lite.
-
Dynamic analysis enhances issue resolution
Embedding dynamic analysis into an LLM repair agent yields a claimed 79.4% resolution rate on SWE-bench Verified while cutting token use by about 25%.
-
ARISE: A Repository-level Graph Representation and Toolset for Agentic Program Repair and Fault Localization
A multi-granularity program graph with intra-procedural def-use edges and a data-flow slicing tool raises SWE-agent resolve rate on SWE-bench Lite by 4.7 points under a fixed open-source backbone.
-
Rethinking the Value of Agent-Generated Tests for LLM-Based Software Engineering Agents
Agent-generated tests mainly act as observational feedback channels and do not meaningfully improve issue resolution success in current LLM software engineering agents.
- ContextSniper: AntTrail's Token-Efficient Code Memory for Repository-Level Program Repair