Retrieval-Oriented Code Representations in Agentic Bug Localization

2026-07-12

AI Summary

This study addresses the challenge of efficiently retrieving code context to support file-level defect localization in warehouse-scale tasks using large language model (LLM) agents. The task is formalized as a retrieval problem, and the work systematically evaluates the effectiveness and overhead of diverse code representations under various retrieval and reranking strategies, positioning code representation as a first-class design consideration in agent-based localization pipelines. The authors propose role-aware summaries, which achieve an optimal trade-off between performance and cost: yielding up to a 40% improvement in Hit@5 over file-path baselines while reducing representation footprint by 10.4–20.9× compared to raw source code. Further gains are realized through multi-representation fusion and LLM-based reranking, providing up to 31.9% and 42.0% relative improvements, respectively, culminating in a 94% Hit@6 within the Agentless framework.

Abstract

LLM-based agents are increasingly being used to support software development, yet their performance in repository-level tasks depends on retrieving the right code context. Existing studies have explored file-level localization using traditional information retrieval over file paths and raw source code. However, the role of textual code representations in retrieval and localization remains underexplored. We study file-level bug localization as a representation-driven retrieval problem. Across the Long Code Arena (LCA) and SWE-bench Verified (SWE) datasets, we compare five code representations: file paths, raw source code, and three LLM-generated textual representations. Our experiments include lexical, semantic, and LLM-based retrieval, followed by LLM-based post-retrieval ranking. We quantify the cost incurred by a representation through the representation footprint. We find that the choice of code representation affects both localization effectiveness and cost. Role-aware summaries outperform file-path representations by up to 40% Hit@5 while requiring a representation footprint 10.4 to 20.9x smaller than raw source code. Combining complementary representation results and ranking retrieved candidates with an LLM provides further gains of up to 31.9% and 42.0%, respectively. Overall, role-aware summaries provide the best cost-effectiveness trade-off, while raw source code offers effectiveness in some settings at a significantly higher cost. A case study with Agentless reveals the utility of our techniques within a well-known pipeline, reaching 94% Hit@6 on file localization (+4.7% against the baseline). Our findings suggest that code representation should be treated as a first-class design choice in agentic localization pipelines, guided by pipeline stage and cost-accuracy requirements.

Problem

• bug localization
• code representation
• retrieval
• LLM-based agents
• software development

Innovation

• code representation
• bug localization
• retrieval-oriented
• role-aware summary
• LLM-based agent