一句话：本文系统梳理 ICSE 上基于大语言模型的软件工程研究，反思其基准测试严谨性、数据污染、可复现性与可持续性挑战并提出改进建议。
随着大语言模型（LLM）进入软件工程（SE）研究，基准测试严谨性、数据污染、可复现性以及财务与环境可持续性等新挑战日益突出。本文呼吁研究社区反思这些挑战在当前 SE 研究中的处理方式，并通过对 ICSE 上现有 LLM 驱动 SE 工作的结构化梳理，总结出值得鼓励的实践与持续存在的不足。基于此，作者提出一系列建议：加强基准测试的严谨性、提升研究的可复现性，并正视和降低 LLM 驱动 SE 研究的财务与环境成本。
Software Engineering (SE) research involving the use of Large Language Models (LLMs) has introduced several new challenges related to rigour in benchmarking, contamination, replicability, and sustainability. In this paper, we invite the research community to reflect on how these challenges are addressed in SE. Our results provide a structured overview of current LLM-based SE research at ICSE, highlighting both encouraging practices and persistent shortcomings. We conclude with recommendations to strengthen benchmarking rigour, improve replicability, and address the financial and environmental costs of LLM-based SE.
DOI 原文 · arXiv · PDF（开放获取） · DBLP
@inproceedings{WilliamsKHAPS26,
title = {Empirical and Sustainability Aspects of Software Engineering Research in the Era of Large Language Models: A Reflection},
author = {David Williams and Maria Kechagia and Max Hort and Aldeida Aleti and Justyna Petke and Federica Sarro},
booktitle = {ICSE},
year = {2026},
doi = {10.1145/3786582.3786827}
}