SR-Eval: Evaluating LLMs on Code Generation under Stepwise Requirement Refinement

• URL: http://arxiv.org/abs/2509.18808v1
• Date: Tue, 23 Sep 2025 08:59:05 GMT
• Title: SR-Eval: Evaluating LLMs on Code Generation under Stepwise Requirement Refinement
• Abstract summary: Large language models (LLMs) have achieved remarkable progress in code generation.<n>This mismatch limits the understanding of how well LLMs can support real-world development.<n>We present SR-Eval, a benchmark specifically designed to assess LLMs on iterative code generation under Stepwise requirements Refinement.
• Score: 10.05571095209475
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large language models (LLMs) have achieved remarkable progress in code generation. However, existing benchmarks mainly formalize the task as a static, single-turn problem, overlooking the stepwise requirement changes and iterative workflows in real-world software development. This mismatch limits the understanding of how well LLMs can support real-world development workflows. Constructing such iterative benchmarks is challenging due to the lack of public interaction traces and the difficulty of creating discriminative, turn-specific test cases. To bridge this gap, we present SR-Eval, a benchmark specifically designed to assess LLMs on iterative code generation under Stepwise requirements Refinement. SR-Eval spans both function-level and repository-level tasks in Python and Java, enabling fine-grained and progressive evaluation across evolving requirements. The construction of SR-Eval follows a carefully designed pipeline that first leverages a multi-agent-based requirement generation method to simulate the development process and recover the multi-round interaction process from final requirements, then employs a semantic-aware discriminative test case generation component to ensure discriminative and consistent evaluation at each turn. SR-Eval comprises 443 multi-turn tasks and 1,857 questions at both function and repository levels. Using SR-Eval, we evaluate 11 representative LLMs with three prompting strategies that simulate different usage patterns. Results show that iterative code generation under stepwise requirement refinement remains highly challenging: the best-performing model achieves only 22.67% completion rate on function-level tasks and 20.00% on repository-level tasks. We further observe that prompting strategies substantially influence performance, highlighting the need for the development of advanced methods.

Related papers

• A2RBench: An Automatic Paradigm for Formally Verifiable Abstract Reasoning Benchmark Generation [59.98516959731531]
  Abstract reasoning ability reflects intelligence and capacity of LLMs to extract and apply abstract rules.<n>Existing benchmarks either rely on expensive manual annotation, limiting their scale, or risk measuring memorization rather than genuine reasoning.<n>We introduce an automated pipeline named A2RBench, encompassing generation, expansion, evaluation, and analysis.
  arXiv  Detail & Related papers  (2026-05-17T06:14:20Z)
• REBENCH: A Procedural, Fair-by-Construction Benchmark for LLMs on Stripped-Binary Types and Names (Extended Version) [32.69257865315466]
  Large Language Models (LLMs) have achieved remarkable progress in recent years, driving their adoption across a wide range of domains.<n>In reverse engineering, LLMs are increasingly applied to critical tasks such as function and variable name recovery and type inference.<n>Existing studies rely on disparate datasets, preprocessing pipelines, and evaluation metrics.<n>We present REBench, a comprehensive benchmark dataset for evaluating LLMs on binary reverse engineering tasks.
  arXiv  Detail & Related papers  (2026-04-30T02:03:37Z)
• RAGCap-Bench: Benchmarking Capabilities of LLMs in Agentic Retrieval Augmented Generation Systems [31.4909149697414]
  Retrieval-Augmented Generation (RAG) mitigates key limitations of Large Language Models (LLMs)<n>Recent work extends this paradigm through agentic RAG systems, where LLMs act as agents to iteratively plan, retrieve, and reason over complex queries.<n>We propose RAGCap-Bench, a capability-oriented benchmark for fine-grained evaluation of intermediate tasks in agentic RAG.
  arXiv  Detail & Related papers  (2025-10-15T04:13:00Z)
• MM-HELIX: Boosting Multimodal Long-Chain Reflective Reasoning with Holistic Platform and Adaptive Hybrid Policy Optimization [103.74675519953898]
  Long-chain reflective reasoning is a prerequisite for solving complex real-world problems.<n>We build a benchmark consisting 1,260 samples of 42 challenging synthetic tasks.<n>We generate post-training data and explore learning paradigms for exploiting such data.
  arXiv  Detail & Related papers  (2025-10-09T17:53:58Z)
• Towards Self-Evolving Benchmarks: Synthesizing Agent Trajectories via Test-Time Exploration under Validate-by-Reproduce Paradigm [60.36837655498119]
  We propose a Trajectory-based validated-by-Reproducing Agent-benchmark Complexity Evolution framework.<n>This framework takes an original task from an existing benchmark and encourages agents to evolve it into a new task with higher difficulty.<n>Experiments on the GAIA benchmark demonstrate that the TRACE framework consistently enhances task complexity while improving the reliability of correctness.
  arXiv  Detail & Related papers  (2025-10-01T01:52:52Z)
• Dynamic Benchmark Construction for Evaluating Large Language Models on Real-World Codes [33.80591142965565]
  We present CODE2BENCH, a pipeline for dynamically constructing robust and contamination-resistant benchmarks from real-world GitHub repositories.<n>Specifically, CODE2BENCH introduces three key innovations: (1) Automated Dynamism, achieved through periodic ingestion of recent code to minimize training data contamination; (2) Scope Graph-based dependency analysis, which enables structured classification of functions into benchmark instances with controlled dependency levels; and (3) Property-Based Testing (PBT) for the automated synthesis of rigorous test suites.
  arXiv  Detail & Related papers  (2025-08-10T05:06:36Z)
• EIFBENCH: Extremely Complex Instruction Following Benchmark for Large Language Models [64.70546873396624]
  We present the Extremely Complex Instruction Following Benchmark (EIFBENCH) for evaluating large language models (LLMs)<n>EIFBENCH includes multi-task scenarios that enable comprehensive assessment across diverse task types concurrently.<n>We also propose the Segment Policy Optimization (SegPO) algorithm to enhance the LLM's ability to accurately fulfill multi-task workflow.
  arXiv  Detail & Related papers  (2025-06-10T02:39:55Z)
• How Effective are Generative Large Language Models in Performing Requirements Classification? [4.429729688079712]
  This study explores the effectiveness of three generative large language models (LLMs) performing both binary and multi-class requirements classification.<n>Our study concludes that while factors like prompt design and LLM architecture are universally important, others-such as dataset variations-have a more situational impact, depending on the complexity of the classification task.
  arXiv  Detail & Related papers  (2025-04-23T14:41:11Z)
• TuRTLe: A Unified Evaluation of LLMs for RTL Generation [0.6010802600885173]
  We propose TuRTLe, a unified evaluation framework designed to assess LLMs across key RTL generation tasks.<n>We benchmark a diverse set of open LLMs and analyze their strengths and weaknesses in EDA-specific tasks.<n>Our results show that reasoning-based models, such as DeepSeek R1, consistently outperform others across multiple evaluation criteria.
  arXiv  Detail & Related papers  (2025-03-31T07:43:12Z)
• Continual LLaVA: Continual Instruction Tuning in Large Vision-Language Models [93.5327725085853]
  Continual LLaVA is a rehearsal-free method tailored for continual instruction tuning in LVLMs. Experiments indicate that the proposed Continual LLaVA outperforms previous methods by significantly reducing the forgetting during the continual instruction tuning process.
  arXiv  Detail & Related papers  (2024-11-04T19:55:32Z)
• COrAL: Order-Agnostic Language Modeling for Efficient Iterative Refinement [80.18490952057125]
  Iterative refinement has emerged as an effective paradigm for enhancing the capabilities of large language models (LLMs) on complex tasks. We propose Context-Wise Order-Agnostic Language Modeling (COrAL) to overcome these challenges. Our approach models multiple token dependencies within manageable context windows, enabling the model to perform iterative refinement internally.
  arXiv  Detail & Related papers  (2024-10-12T23:56:19Z)
• Benchmarking Agentic Workflow Generation [80.74757493266057]
  We introduce WorfBench, a unified workflow generation benchmark with multi-faceted scenarios and intricate graph workflow structures.<n>We also present WorfEval, a systemic evaluation protocol utilizing subsequence and subgraph matching algorithms.<n>We observe that the generated can enhance downstream tasks, enabling them to achieve superior performance with less time during inference.
  arXiv  Detail & Related papers  (2024-10-10T12:41:19Z)
• Can Long-Context Language Models Subsume Retrieval, RAG, SQL, and More? [54.667202878390526]
  Long-context language models (LCLMs) have the potential to revolutionize our approach to tasks traditionally reliant on external tools like retrieval systems or databases. We introduce LOFT, a benchmark of real-world tasks requiring context up to millions of tokens designed to evaluate LCLMs' performance on in-context retrieval and reasoning. Our findings reveal LCLMs' surprising ability to rival state-of-the-art retrieval and RAG systems, despite never having been explicitly trained for these tasks.
  arXiv  Detail & Related papers  (2024-06-19T00:28:58Z)

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.