提出分层并行代理框架，解决网页信息搜索中的数据过载与延迟问题。
InfoSeeker: A Scalable Hierarchical Parallel Agent Framework for Web Information Seeking
- 分层架构：主机-管理者-工作者协同，实现任务分解与并行处理
- 速度提升3-5倍，中文浏览任务准确率达52.9%
- 适合大规模多源信息整合场景，尤其适用于长流程推理任务
近期的智能体搜索系统在深度多步推理方面取得显著进展，但往往忽视了大规模信息综合带来的挑战，即需从多个异构来源聚合大量证据。因此，现有大语言模型代理系统在数据密集型场景中面临严重限制，包括上下文饱和、错误传播累积和高端到端延迟。为此，我们提出 ramework，一个基于近可分解性原则的分层框架，包含战略级的 extit{Host}、多个 extit{Managers}和并行的 extit{Workers}。通过在管理者层级引入聚合与反思机制，框架强制实现上下文隔离，有效防止饱和与错误传播；同时，工作层的并行性显著加快整体任务执行速度，缓解显著延迟。在两个互补基准上的评估表明，该框架兼具高效性（3-5×加速）与有效性，在WideSearch-en上达到8.4%的成功率，在BrowseComp-zh上实现52.9%的准确率。代码已开源于https://github.com/agent-on-the-fly/InfoSeeker。
原文摘要 · Abstract (English)
Recent agentic search systems have made substantial progress by emphasising deep, multi-step reasoning. However, this focus often overlooks the challenges of wide-scale information synthesis, where agents must aggregate large volumes of heterogeneous evidence across many sources. As a result, most existing large language model agent systems face severe limitations in data-intensive settings, including context saturation, cascading error propagation, and high end-to-end latency. To address these challenges, we present \framework, a hierarchical framework based on principle of near-decomposability, containing a strategic \textit{Host}, multiple \textit{Managers} and parallel \textit{Workers}. By leveraging aggregation and reflection mechanisms at the Manager layer, our framework enforces strict context isolation to prevent saturation and error propagation. Simultaneously, the parallelism in worker layer accelerates the speed of overall task execution, mitigating the significant latency. Our evaluation on two complementary benchmarks demonstrates both efficiency ($ 3-5 \times$ speed-up) and effectiveness, achieving a $8.4\%$ success rate on WideSearch-en and $52.9\%$ accuracy on BrowseComp-zh. The code is released at https://github.com/agent-on-the-fly/InfoSeeker
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。