OpenDS-World：真实开放网络环境下的深度搜索基准

AI 摘要

研究者提出 OpenDS 任务与 OpenDS-World 基准，用于评估自主多步深度搜索智能体在真实开放网络环境中的端到端研究能力。该基准包含 200 个跨模态、跨格式且需交互访问的挑战性任务，并经人工验证确保证据有效与答案唯一。在统一工具环境下评测 12 个前沿模型，最佳模型准确率为 39.0%，而人类在 1000 次独立尝试中的准确率为 21.1%。分析还揭示了不同模型在研究策略、工具协调与长程搜索管理上的显著差异与持续挑战。

正文节选

Deep Search in Real-World Web Environments Abstract OpenDS-World introduces a live-web benchmark to evaluate autonomous multi-step deep search agents, revealing significant gaps between frontier models and human performance. Despite substantial progress in Deep Search agents, their ability to conduct end-to-end research in real-world open Web environments remains insufficiently evaluated. Unlike static retrieval settings, information on the real-world Web is distributed across diverse sources, m

发布时间：—

抓取时间：2026-10-23 12:42

来源机构：Hugging Face