Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2604.02729 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908934657277952 |
|---|---|
| author | Zeng, Puyu Wang, Zhaoxi Duan, Zhixu Feng, Liang Wang, Shaobo Wang, Cunxiang Wang, Jinghang Zhao, Bing Wei, Hu Zhang, Linfeng |
| author_facet | Zeng, Puyu Wang, Zhaoxi Duan, Zhixu Feng, Liang Wang, Shaobo Wang, Cunxiang Wang, Jinghang Zhao, Bing Wei, Hu Zhang, Linfeng |
| contents | Code generation and comprehension by Large Language Models (LLMs) have emerged as core drivers of industrial intelligence and decision optimization, finding widespread application in fields such as finance, automation, and aerospace. Although recent advancements have demonstrated the remarkable potential of LLMs in general code generation, existing benchmarks are mainly confined to single domains and languages. Consequently, they fail to effectively evaluate the generalization capabilities required for real-world industrial applications or to reflect the coding proficiency demanded by complex industrial scenarios. To bridge this gap, we introduce IndustryCode, the first comprehensive benchmark designed to span multiple industrial domains and programming languages. IndustryCode comprises 579 sub-problems derived from 125 primary industrial challenges, accompanied by rigorous problem descriptions and test cases. It covers a wide range of fields, including finance, automation, aerospace, and remote sensing-and incorporates diverse programming languages such as MATLAB, Python, C++, and Stata. In our evaluation, the top-performing model, Claude 4.5 Opus, achieved an overall accuracy of 68.1% on sub-problems and 42.5% main problems. The benchmark dataset and automated evaluation code will be made publicly available upon acceptance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_02729 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | IndustryCode: A Benchmark for Industry Code Generation Zeng, Puyu Wang, Zhaoxi Duan, Zhixu Feng, Liang Wang, Shaobo Wang, Cunxiang Wang, Jinghang Zhao, Bing Wei, Hu Zhang, Linfeng Software Engineering Artificial Intelligence Computation and Language Code generation and comprehension by Large Language Models (LLMs) have emerged as core drivers of industrial intelligence and decision optimization, finding widespread application in fields such as finance, automation, and aerospace. Although recent advancements have demonstrated the remarkable potential of LLMs in general code generation, existing benchmarks are mainly confined to single domains and languages. Consequently, they fail to effectively evaluate the generalization capabilities required for real-world industrial applications or to reflect the coding proficiency demanded by complex industrial scenarios. To bridge this gap, we introduce IndustryCode, the first comprehensive benchmark designed to span multiple industrial domains and programming languages. IndustryCode comprises 579 sub-problems derived from 125 primary industrial challenges, accompanied by rigorous problem descriptions and test cases. It covers a wide range of fields, including finance, automation, aerospace, and remote sensing-and incorporates diverse programming languages such as MATLAB, Python, C++, and Stata. In our evaluation, the top-performing model, Claude 4.5 Opus, achieved an overall accuracy of 68.1% on sub-problems and 42.5% main problems. The benchmark dataset and automated evaluation code will be made publicly available upon acceptance. |
| title | IndustryCode: A Benchmark for Industry Code Generation |
| topic | Software Engineering Artificial Intelligence Computation and Language |
| url | https://arxiv.org/abs/2604.02729 |