Self-Bootstrapping Automated Program Repair: Using LLMs to Generate and Evaluate Synthetic Training Data for Bug Repair

TL;DR

This paper introduces a self-bootstrapping approach using LLMs to generate and evaluate synthetic training data, significantly improving automated program repair across multiple languages and bug types.

Contribution

It presents a novel two-phase methodology for synthetic data generation and quality assessment, enhancing APR performance with less computational effort.

Findings

Synthetic dataset improved Top@1 prediction accuracy by 47%

Achieved statistically significant improvements over baseline systems

Validated approach across 12 programming languages and 13 bug categories

Abstract

This paper presents a novel methodology for enhancing Automated Program Repair (APR) through synthetic data generation utilizing Large Language Models (LLMs). Current APR systems are constrained by the limited availability of high-quality training data encompassing diverse bug types across multiple programming languages. The proposed approach addresses this limitation through a two-phase process: a synthetic sample generation followed by a rigorous quality assessment. Multiple state-of-the-art LLMs were employed to generate approximately 30,000 paired examples of buggy and fixed code across 12 programming languages and 13 bug categories. Subsequently, these samples underwent cross-model evaluation against five criteria: correctness, code quality, security, performance, and completeness. Experimental evaluation on the VulRepair test set dataset showed statistically significant improvements in Perfect Prediction rates, with the quality-filtered synthetic dataset achieving 17.18% (Top@1) and 23.00% (Top@5) compared to the baseline's 11.68% and 18.88% respectively, representing a 47% relative improvement in Top@1 and 22% in Top@5. The methodology was validated through rigorous statistical testing, including ANOVA and post-hoc Tukey's Honest Significant Difference analysis. Furthermore, the best-performing configurations surpassed existing systems despite using a less computationally intensive decoding strategy. This research establishes a self-bootstrapping paradigm in which LLMs generate and evaluate their own training data, suggesting promising directions for addressing data scarcity in similar software engineering tasks and advancing the development of robust, adaptable tools for automated code maintenance.