arXiv cs.AI· David Nukrai, Bar Alon, Nativ Levy, Ido Andrew Atad, Ido Cohen, Itamar Zimerman, Lior Wolf, Eli David, Ilan Naiman·· 3 天前AI 评分34
EvoWeb：面向持续演化 Web 应用的编码智能体基准
EvoWeb: A Benchmark for Coding Agents on Evolving Web Applications
AI 导读
研究者推出 EvoWeb 基准，包含覆盖 32 个 Web 应用的 320 条累积请求，按功能依赖图组织，由多模态 Web 智能体在真实浏览器中逐轮测试新功能、跨请求集成与既有功能。实验显示开源与闭源前沿模型表现差异明显，跨请求集成是主要难点，依赖更深或更多的后续请求解决准确率更低，运行时错误也随请求推进而增多。
正文 · 原文
Abstract:Web application development often proceeds through successive requests that extend an existing implementation. Yet web-generation benchmarks primarily evaluate development toward fixed specifications. We introduce EvoWeb, a novel benchmark comprising 320 cumulative requests over 32 web applications, grounded in real-world product features and structured by functional dependency graphs. Agents iteratively modify a persistent codebase, while a multimodal web agent tests each intermediate application in a live browser, assessing newly requested functionality, cross-request integration, and previously requested functionality. To characterize reliability throughout development, we integrate trajectory performance across user-tolerance thresholds rather than impose a fixed stopping criterion. Experiments across open- and closed-source frontier models show that EvoWeb reveals clear performance differences between models, with cross-request integration emerging as a key challenge. Further analysis shows that later requests with deeper or more numerous dependencies are solved less accurately, while runtime errors become more frequent over successive requests. EvoWeb provides a testbed for evaluating and improving coding agents in long-horizon, multi-request development.
| Comments: | 24 pages, 6 figures. Code: this https URL ; Dataset: this https URL |
| Subjects: | Artificial Intelligence (cs.AI); Software Engineering (cs.SE) |
| Cite as: | arXiv:2610.16402 [cs.AI] |
| (or arXiv:2610.16402v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.16402 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Bar Alon [view email]
[v1]
Mon, 12 Oct 2026 08:42:42 UTC (679 KB)
来源：arXiv cs.AI · arxiv.org