RepoGenReflex: Enhancing Repository-Level Code Completion with Verbal Reinforcement and Retrieval-Augmented Generation

TL;DR

RepoGenReflex is a novel framework that improves repository-level code completion by dynamically selecting optimal results through Retrieval-Augmented Generation combined with Verbal Reinforcement Learning, validated on a new real-world benchmark.

Contribution

It introduces RepoGenReflex, a dynamic, generic framework that enhances code completion accuracy by integrating RAG and VRL, along with a new benchmark RepoGenEval for evaluation.

Findings

Significant accuracy improvements over baseline methods.

Consistent performance across diverse code completion scenarios.

Enhanced relevance and robustness in code suggestions.

Abstract

In real-world software engineering tasks, solving a problem often requires understanding and modifying multiple functions, classes, and files across a large codebase. Therefore, on the repository level, it is crucial to extract the relevant information to achieve accurate code completion effectively. Existing code completion tools have achieved some success, but they struggle to optimize the retrieval and generation process dynamically. In this paper, we propose RepoGenReflex, a generic, dynamic, effective framework to address this challenge. By leveraging the Retrieval-Augmented Generation (RAG) enhanced with Verbal Reinforcement Learning (VRL), it can dynamically choose the optimal results for repository-level code completion. RepoGenReflex uses Reflector to give directional feedback to the next loop. RepoGenReflex chooses the optimal results stored in the Experience cache based on the RAG-VRL loop. To validate the framework's generalization ability, we propose a new benchmark RepoGenEval, which consists of the latest, high-quality real-world repositories in line completion scenarios. Our experiments demonstrate that RepoGenReflex achieves significant improvements after optimizing the Reflector component, resulting in enhanced accuracy and relevance of code completions. Additionally, RepoGenReflex consistently demonstrates superior performance and effectiveness across standard code completion tasks, highlighting the robustness and adaptability of our framework.