AACR-Bench: Evaluating Automatic Code Review with Holistic Repository-Level Context
This work addresses the problem of inaccurate and limited evaluation for Automated Code Review systems, particularly for researchers and developers, and is incremental as it builds upon existing benchmarks by improving data quality and scope.
The paper tackles the limitations of existing benchmarks for evaluating Large Language Models in Automated Code Review by introducing AACR-Bench, which provides multi-language, repository-level context and uses an AI-assisted, expert-verified annotation pipeline, resulting in a 285% increase in defect coverage and revealing that context granularity and retrieval methods significantly impact performance.
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 .