New framework Loc2Repair enhances LLM code repair effectiveness

A new framework called Loc2Repair has been developed to evaluate the impact of file-level issue localization in repository-level LLM repair. This framework decouples localization and repair, allowing for controlled analysis of different components. Experiments using three repair backbones on SWE-bench Verified demonstrated that explicit file-level localization consistently improves the resolution rate and reduces the mean elapsed time for LLM-based code repair. AI

IMPACT This framework could lead to more effective and efficient LLM-based code repair tools.

RANK_REASON The cluster contains a research paper detailing a new framework for evaluating LLM capabilities. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• Loc2Repair
• Mohammad Nour Al Awad
• SWE-bench Verified

• paper
• product

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →