MRG-Bench: Evaluating and Exploring the Requirements of Context for Repository-Level Code Generation

• URL: http://arxiv.org/abs/2508.02998v1
• Date: Tue, 05 Aug 2025 01:53:45 GMT
• Title: MRG-Bench: Evaluating and Exploring the Requirements of Context for Repository-Level Code Generation
• Authors: Haiyang Li,
• Abstract summary: We introduce textbfMRG-Bench, a novel dataset that provides a more accurate evaluation of large language models.<n>We conduct experiments including large language models, long-context models, and RAG-related methods.<n>Results show that the majority of methods suffer from "textbfdifficulty in understanding user requirements," failing to comprehend their assigned tasks accurately.
• Score: 0.7342677574855649
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Large Language Models (LLMs) have demonstrated impressive capabilities in code generation. However, current evaluation datasets suffer from issues such as the lack of runnable test cases, deviation from the distribution of real-world code, and the ability to evaluate only the Python language. These limitations undermine the credibility of the evaluation results. To address these limitations, we introduce \textbf{MRG-Bench} (Multi-language Repository-level Code Generation Benchmark), a novel dataset that provides a more accurate evaluation of LLMs in practical repository-level code generation tasks. MRG-Bench has three main features: (1) practical data sourced from real-world code repositories that align to the practical distribution, (2) multiple programming languages support, including Python, Java, and Go, and (3) project-level runnable test cases to assess the quality of the generated code. Based on MRG-Bench, we conducted extensive experiments including large language models, long-context models, and RAG-related methods. These evaluation results demonstrate that \textbf{current repository-level code generation techniques suffer from significant performance deficiencies}. To further investigate why models fail, we designed novel experiments to annotate the underlying causes of generation errors. The results explicitly show that the majority of methods suffer from "\textbf{difficulty in understanding user requirements}," failing to comprehend their assigned tasks accurately. Moreover, the impact of different repository-level contexts on this issue exhibits significant disparities across different programming languages, suggesting that, in practice, specialized contextual information needs to be designed for different languages.

Related papers

• IFEvalCode: Controlled Code Generation [69.28317223249358]
  The paper introduces forward and backward constraints generation to improve the instruction-following capabilities of Code LLMs.<n>The authors present IFEvalCode, a multilingual benchmark comprising 1.6K test samples across seven programming languages.
  arXiv  Detail & Related papers  (2025-07-30T08:08:48Z)
• Turning the Tide: Repository-based Code Reflection [52.13709676656648]
  We introduce LiveRepoReflection, a benchmark for evaluating code understanding and generation in multi-file repository contexts.<n>1,888 rigorously filtered test cases across $6$ programming languages to ensure diversity, correctness, and high difficulty.<n>We also create RepoReflection-Instruct, a large-scale, quality-filtered instruction-tuning dataset derived from diverse sources.
  arXiv  Detail & Related papers  (2025-07-14T02:36:27Z)
• Is Compression Really Linear with Code Intelligence? [60.123628177110206]
  textitFormat Annealing is a lightweight, transparent training methodology designed to assess the intrinsic capabilities of pre-trained models equitably.<n>Our empirical results reveal a fundamental logarithmic relationship between measured code intelligence and bits-per-character (BPC)<n>Our work provides a more nuanced understanding of compression's role in developing code intelligence and contributes a robust evaluation framework in the code domain.
  arXiv  Detail & Related papers  (2025-05-16T16:59:14Z)
• TestBench: Evaluating Class-Level Test Case Generation Capability of Large Language Models [8.22619177301814]
  We introduce TestBench, a benchmark for class-level LLM-based test case generation. We construct a dataset of 108 Java programs from 9 real-world, large-scale projects on GitHub. We propose a fine-grained evaluation framework that considers five aspects of test cases: syntactic correctness, compilation correctness, test correctness, code coverage rate, and defect detection rate.
  arXiv  Detail & Related papers  (2024-09-26T06:18:06Z)
• ComplexCodeEval: A Benchmark for Evaluating Large Code Models on More Complex Code [29.178248778212588]
  ComplexCodeEval is a benchmark designed to assess large language models (LLMs) in various development tasks. It includes 3,897 Java samples and 7,184 Python samples from high-star GitHub repositories.
  arXiv  Detail & Related papers  (2024-09-16T13:43:04Z)
• What's Wrong with Your Code Generated by Large Language Models? An Extensive Study [80.18342600996601]
  Large language models (LLMs) produce code that is shorter yet more complicated as compared to canonical solutions. We develop a taxonomy of bugs for incorrect codes that includes three categories and 12 sub-categories, and analyze the root cause for common bug types. We propose a novel training-free iterative method that introduces self-critique, enabling LLMs to critique and correct their generated code based on bug types and compiler feedback.
  arXiv  Detail & Related papers  (2024-07-08T17:27:17Z)
• CodeRAG-Bench: Can Retrieval Augment Code Generation? [78.37076502395699]
  We conduct a systematic, large-scale analysis of code generation using retrieval-augmented generation.<n>We first curate a comprehensive evaluation benchmark, CodeRAG-Bench, encompassing three categories of code generation tasks.<n>We examine top-performing models on CodeRAG-Bench by providing contexts retrieved from one or multiple sources.
  arXiv  Detail & Related papers  (2024-06-20T16:59:52Z)
• On the Impacts of Contexts on Repository-Level Code Generation [5.641402231731082]
  We present RepoExec, a novel benchmark designed to evaluate repository-level code generation.<n>We focus on three key aspects: executability, functional correctness through comprehensive test case generation, and accurate utilization of cross-file contexts.
  arXiv  Detail & Related papers  (2024-06-17T10:45:22Z)
• Quantifying Contamination in Evaluating Code Generation Capabilities of Language Models [27.24738197172374]
  Large language models have achieved remarkable performance on various code generation benchmarks. There have been growing concerns regarding potential contamination of these benchmarks as they may be leaked into pretraining and finetuning data. We show that there are substantial overlap between popular code generation benchmarks and open training corpus, and models perform significantly better on the subset of the benchmarks where similar solutions are seen during training.
  arXiv  Detail & Related papers  (2024-03-06T21:45:35Z)
• ML-Bench: Evaluating Large Language Models and Agents for Machine Learning Tasks on Repository-Level Code [76.84199699772903]
  ML-Bench is a benchmark rooted in real-world programming applications that leverage existing code repositories to perform tasks. To evaluate both Large Language Models (LLMs) and AI agents, two setups are employed: ML-LLM-Bench for assessing LLMs' text-to-code conversion within a predefined deployment environment, and ML-Agent-Bench for testing autonomous agents in an end-to-end task execution within a Linux sandbox environment.
  arXiv  Detail & Related papers  (2023-11-16T12:03:21Z)
• L2CEval: Evaluating Language-to-Code Generation Capabilities of Large Language Models [102.00201523306986]
  We present L2CEval, a systematic evaluation of the language-to-code generation capabilities of large language models (LLMs) We analyze the factors that potentially affect their performance, such as model size, pretraining data, instruction tuning, and different prompting methods. In addition to assessing model performance, we measure confidence calibration for the models and conduct human evaluations of the output programs.
  arXiv  Detail & Related papers  (2023-09-29T17:57:00Z)
• ICE-Score: Instructing Large Language Models to Evaluate Code [7.556444391696562]
  We propose textttICE-Score, a new evaluation metric via instructing large language models for code assessments. Our metric addresses the limitations of existing approaches by achieving superior correlations with functional correctness and human preferences. Our results demonstrate that our metric surpasses state-of-the-art metrics for code generation.
  arXiv  Detail & Related papers  (2023-04-27T16:38:17Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.