AACR-Bench: Evaluating Automatic Code Review with Holistic Repository-Level Context

TL;DR

AACR-Bench is a new, comprehensive benchmark for evaluating Large Language Models in Automated Code Review, supporting multiple languages and using expert-verified annotations to improve defect detection and assessment accuracy.

Contribution

Introduces AACR-Bench, a multi-language, repository-level benchmark with an AI-assisted, expert-verified annotation pipeline, enhancing defect coverage and evaluation rigor for LLM-based ACR.

Findings

Previous benchmarks had limited language support and noisy ground truth.

Model performance varies significantly with context granularity and retrieval methods.

Evaluation reveals that prior assessments may have misjudged LLM capabilities.

Abstract

High-quality evaluation benchmarks are pivotal for deploying Large Language Models (LLMs) in Automated Code Review (ACR). However, existing benchmarks suffer from two critical limitations: first, the lack of multi-language support in repository-level contexts, which restricts the generalizability of evaluation results; second, the reliance on noisy, incomplete ground truth derived from raw Pull Request (PR) comments, which constrains the scope of issue detection. To address these challenges, we introduce AACR-Bench a comprehensive benchmark that provides full cross-file context across multiple programming languages. Unlike traditional datasets, AACR-Bench employs an "AI-assisted, Expert-verified" annotation pipeline to uncover latent defects often overlooked in original PRs, resulting in a 285% increase in defect coverage. Extensive evaluations of mainstream LLMs on AACR-Bench reveal that previous assessments may have either misjudged or only partially captured model capabilities due to data limitations. Our work establishes a more rigorous standard for ACR evaluation and offers new insights on LLM based ACR, i.e., the granularity/level of context and the choice of retrieval methods significantly impact ACR performance, and this influence varies depending on the LLM, programming language, and the LLM usage paradigm e.g., whether an Agent architecture is employed. The code, data, and other artifacts of our evaluation set are available at https://github.com/alibaba/aacr-bench .