arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.01331cs.MAcs.CL
收藏

A-MapReduce:通过代理MapReduce执行广泛搜索

A-MapReduce: Executing Wide Search via Agentic MapReduce

Beijing Advanced Innovation Center for Future Blockchain and Privacy Computing, School of Artificial Intelligence, Beihang University(北京未来区块链与隐私计算创新中心,人工智能学院,北京航空航天大学);National University of Singapore(新加坡国立大学);Tsinghua University(清华大学);BNRist, Tsinghua University(BNRist,清华大学)

Mingju Chen, Guibin Zhang, Heng Chang, Xiao Wang, Shiji Zhou

AI总结:

A-MapReduce通过引入MapReduce范式,提出一种多代理框架,将广泛搜索转化为水平结构的检索问题,实现高效且低成本的广泛搜索执行。

摘要:

当前基于大型语言模型(LLM)的多代理系统在深度研究任务中表现出系统性优势,这些任务强调迭代性和垂直结构的信息检索。然而,当面对以大规模、广度导向检索为特征的广泛搜索任务时,现有多代理框架,主要围绕顺序性、垂直结构的推理设计,仍陷入扩展搜索目标和低效长周期执行的困境。为弥合这一差距,我们提出了A-MapReduce,一种受MapReduce范式启发的多代理执行框架,将广泛搜索重新表述为水平结构的检索问题。具体而言,A-MapReduce通过任务自适应分解和结构化结果聚合,实现对大规模检索目标的并行处理。同时,它利用经验记忆来驱动查询条件下的任务分配和重组的持续进化,使在大规模广泛搜索环境中实现逐步改进。在五个多代理基准上的广泛实验表明,A-MapReduce在(i)高性能方面表现优异,其在WideSearch和DeepWideSearch上达到最先进的性能,并在使用OpenAI o3或Gemini 2.5 Pro作为基础模型时,平均Item F1提升达5.11% -17.50%;(ii)在成本效益和效率方面表现优异,提供更优越的成本性能权衡,并将运行时间比代表性多代理基线减少45.8%。代码可在https://github.com/mingju-c/AMapReduce获取。

英文摘要:

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 https://github.com/mingju-c/AMapReduce.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑