Reformulate, Retrieve, Localize: Agents for Repository-Level Bug Localization

2025-12-07

AI Summary

To address low defect localization accuracy in large-scale software repositories—where conventional information retrieval–based methods suffer from noise in bug reports—this paper proposes an LLM-driven agent framework that synergistically optimizes the BM25 retrieval pipeline via lightweight query reformulation and key summary extraction. Our approach requires no fine-tuning of open-source large language models, yet achieves semantic-enriched query rewriting and context-aware information distillation. It constructs an end-to-end automated localization pipeline supporting repository-level, fine-grained file localization. Experimental results demonstrate a 35% improvement in Mean Reciprocal Rank (MRR) over the BM25 baseline and up to a 22% higher Top-10 file recall rate compared to SWE-agent. The method significantly enhances both localization efficiency and robustness against noisy defect reports.

Technology Category

• Search and Optimization: Local Search
• Machine Learning: Large Multimodal Models (LMMs)
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Bug localization remains a critical yet time-consuming challenge in large-scale software repositories. Traditional information retrieval-based bug localization (IRBL) methods rely on unchanged bug descriptions, which often contain noisy information, leading to poor retrieval accuracy. Recent advances in large language models (LLMs) have improved bug localization through query reformulation, yet the effect on agent performance remains unexplored. In this study, we investigate how an LLM-powered agent can improve file-level bug localization via lightweight query reformulation and summarization. We first employ an open-source, non-fine-tuned LLM to extract key information from bug reports, such as identifiers and code snippets, and reformulate queries pre-retrieval. Our agent then orchestrates BM25 retrieval using these preprocessed queries, automating localization workflow at scale. Using the best-performing query reformulation technique, our agent achieves 35% better ranking in first-file retrieval than our BM25 baseline and up to +22% file retrieval performance over SWE-agent.

Problem

• Improves bug localization via LLM query reformulation
• Automates retrieval workflow for large software repositories
• Enhances file-level accuracy over traditional IRBL methods

Innovation

• LLM extracts key bug report details for query reformulation
• Agent orchestrates BM25 retrieval with preprocessed queries
• Lightweight reformulation improves first-file retrieval ranking by 35%