一个名为Ko-WideSearch的新基准已被开发出来，用于评估网络代理的广度搜索能力，重点关注穷举集枚举而非基于深度的问答。这个韩语基准通过自动化流程构建，包含190个实体和16个类别的228张表格。对20个网络代理的初步测试显示，即使在整体集合成员资格被正确识别的情况下，它们在准确恢复行级属性方面也持续失败，这表明当前AI系统面临重大挑战。 AI
影响 突出了网络代理能力的一个关键差距，表明当前模型在结构化数据提取和穷举枚举方面需要改进。
排序理由 该集群描述了一个用于评估AI能力的新学术基准，已在arXiv上发布。
在 Hugging Face Daily Papers 阅读 →
AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →
arXiv:2606.27595v1 Announce Type: new Abstract: Web-agent benchmarks overwhelmingly measure depth -- pinning one obscure answer behind a chain of constraints -- while breadth, exhaustively enumerating a closed set and filling each item's attributes, is barely evaluated, especiall…
Web-agent benchmarks overwhelmingly measure depth -- pinning one obscure answer behind a chain of constraints -- while breadth, exhaustively enumerating a closed set and filling each item's attributes, is barely evaluated, especially outside English. Breadth is also hard to build…
A Korean web-agent benchmark evaluates breadth of search capabilities by requiring complete enumeration of entity memberships with attribute tables, revealing consistent failures in row recovery despite accurate set identification.