Web2BigTable：一個用於網路規模資訊搜尋與擷取的雙層多代理大型語言模型系統

Web2BigTable: A Bi-Level Multi-Agent LLM System for Internet-Scale Information Search and Extraction

April 29, 2026

作者: Yuxuan Huang, Yihang Chen, Zhiyuan He, Yuxiang Chen, Ka Yiu Lee, Huichi Zhou, Weilin Luo, Meng Fang, Jun Wang

cs.AI

摘要

基於代理的網路搜尋正面臨兩大需求：針對單一目標的深度推理，以及跨越多個實體與異構來源的結構化聚合。現有系統在這兩方面皆存在不足：廣度導向任務需要具備廣泛覆蓋度且符合架構規範的輸出，同時保持跨實體一致性；而深度導向任務則要求對長鏈條、多分支的搜尋軌跡進行連貫推理。我們提出Web2BigTable——一個支援雙重模式的網路到表格搜尋多代理框架。該框架採用雙層架構：上層協調器將任務分解為子問題，下層工作代理則並行處理這些子問題。透過「執行—驗證—反思」的閉環流程，框架藉由持久化、人類可讀的外部記憶體，持續優化任務分解與執行效能，並實現單一代理的自我演化更新。執行過程中，工作代理透過共享工作區協調，使部分發現結果可視化，從而減少重複探索、調和衝突證據，並動態適應新出現的覆蓋缺口。Web2BigTable在WideSearch基準上實現突破性表現：Avg@4成功率達38.50（超越第二名7.5倍，其成績為5.10），行F1值達63.53（較第二名提升25.03），項目F1值達80.12（較第二名提升14.42）。該框架同樣能泛化至XBench-DeepSearch的深度導向搜尋任務，達成73.0%的準確率。程式碼已開源於：https://github.com/web2bigtable/web2bigtable。

English

Agentic web search increasingly faces two distinct demands: deep reasoning over a single target, and structured aggregation across many entities and heterogeneous sources. Current systems struggle on both fronts. Breadth-oriented tasks demand schema-aligned outputs with wide coverage and cross-entity consistency, while depth-oriented tasks require coherent reasoning over long, branching search trajectories. We introduce Web2BigTable, a multi-agent framework for web-to-table search that supports both regimes. Web2BigTable adopts a bi-level architecture in which an upper-level orchestrator decomposes the task into sub-problems and lower-level worker agents solve them in parallel. Through a closed-loop run--verify--reflect process, the framework jointly improves decomposition and execution over time via persistent, human-readable external memory, with self-evolving updates to each single-agent. During execution, workers coordinate through a shared workspace that makes partial findings visible, allowing them to reduce redundant exploration, reconcile conflicting evidence, and adapt to emerging coverage gaps. Web2BigTable sets a new state of the art on WideSearch, reaching an Avg@4 Success Rate of 38.50 (7.5times the second best at 5.10), Row F1 of 63.53 (+25.03 over the second best), and Item F1 of 80.12 (+14.42 over the second best). It also generalises to depth-oriented search on XBench-DeepSearch, achieving 73.0 accuracy. Code is available at https://github.com/web2bigtable/web2bigtable.