A-MapReduce：用智能体 MapReduce 执行广度搜索

View PDF HTML (experimental)

Abstract:Contemporary large language model (LLM)-based multi-agent systems exhibit systematic advantages in deep research tasks, which emphasize iterative, vertically structured information seeking. However, when confronted with wide search tasks characterized by large-scale, breadth-oriented retrieval, existing agentic frameworks, primarily designed around sequential, vertically structured reasoning, remain stuck in expansive search objectives and inefficient long-horizon execution. To bridge this gap, we propose A-MapReduce, a MapReduce paradigm-inspired multi-agent execution framework that recasts wide search as a horizontally structured retrieval problem. Concretely, A-MapReduce implements parallel processing of massive retrieval targets through task-adaptive decomposition and structured result aggregation. Meanwhile, it leverages experiential memory to drive the continual evolution of query-conditioned task allocation and recomposition, enabling progressive improvement in large-scale wide-search regimes. Extensive experiments on five agentic benchmarks demonstrate that A-MapReduce is (i) high-performing, achieving state-of-the-art performance on WideSearch and DeepWideSearch, and delivering 5.11% - 17.50% average Item F1 improvements compared with strong baselines with OpenAI o3 or Gemini 2.5 Pro backbones; (ii) cost-effective and efficient, delivering superior cost-performance trade-offs and reducing running time by 45.8\% compared to representative multi-agent baselines. The code is available at this https URL.

Comments:	27 pages
Subjects:	Multiagent Systems (cs.MA); Computation and Language (cs.CL)
Cite as:	arXiv:2602.01331 [cs.MA]
	(or arXiv:2602.01331v2 [cs.MA] for this version)
	https://doi.org/10.48550/arXiv.2602.01331 arXiv-issued DOI via DataCite

Submission history

From: Mingju Chen [view email]
[v1] Sun, 1 Feb 2026 16:53:29 UTC (7,122 KB)
[v2] Fri, 23 Oct 2026 15:19:43 UTC (7,120 KB)