arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2508.01005cs.CLcs.IR
收藏

MAO-ARAG:面向自适应检索增强生成的多智能体编排

MAO-ARAG: Multi-Agent Orchestration for Adaptive Retrieval-Augmented Generation

Renmin University of China(中国人民大学);Baidu Inc.(百度公司)

Yiqun Chen, Erhan Zhang, Lingyong Yan, Shuaiqiang Wang, Jizhou Huang, Dawei Yin, Jiaxin Mao

AI总结:

针对固定RAG流水线难以平衡性能与成本的问题,提出MAO-ARAG多智能体编排框架,通过规划智能体动态选择执行智能体并强化学习优化,实现高质量低成本的问答。

摘要:

在问答(QA)系统中,检索增强生成(RAG)已成为提升回答准确性和减少幻觉问题的关键。RAG系统的架构差异显著,涵盖单轮RAG、迭代RAG和推理RAG,每种架构针对不同类型的查询而设计。由于现实世界查询的复杂性各不相同,固定的RAG流水线往往难以在不同查询间平衡性能和成本效率。为应对这一挑战,我们提出了一种名为MAO-ARAG的自适应RAG框架,该框架利用多智能体编排。我们的自适应RAG被设计为多轮框架。具体而言,我们定义了多个执行智能体,代表典型的RAG模块,如查询重构智能体、文档选择智能体和生成智能体。一个规划智能体智能地从这些执行智能体中选择并整合合适的智能体,为每个查询定制合适的工作流,力求在保持合理成本的同时提供高质量的回答。在每一轮中,规划智能体使用强化学习进行训练,由基于结果的奖励(F1分数)和基于成本的惩罚引导,持续提升回答质量,同时将成本控制在合理范围内。在多个问答数据集上进行的实验表明,我们的方法为每个查询动态规划工作流,不仅实现了高回答质量,还将成本和延迟维持在可接受的范围内。MAO-ARAG的代码位于https://github.com/chenyiqun/Agentic-RAG。

英文摘要:

In question-answering (QA) systems, Retrieval-Augmented Generation (RAG) has become pivotal in enhancing response accuracy and reducing hallucination issues. The architecture of RAG systems varies significantly, encompassing single-round RAG, iterative RAG, and reasoning RAG, each tailored to address different types of queries. Due to the varying complexity of real-world queries, a fixed RAG pipeline often struggles to balance performance and cost efficiency across different queries. To address this challenge, we propose an adaptive RAG framework called MAO-ARAG, which leverages multi-agent orchestration. Our adaptive RAG is conceived as a multi-turn framework. Specifically, we define multiple executor agents, representing typical RAG modules such as query reformulation agents, document selection agent, and generation agents. A planner agent intelligently selects and integrates the appropriate agents from these executors into a suitable workflow tailored for each query, striving for high-quality answers while maintaining reasonable costs. During each turn, the planner agent is trained using reinforcement learning, guided by an outcome-based reward (F1 score) and a cost-based penalty, continuously improving answer quality while keeping costs within a reasonable range. Experiments conducted on multiple QA datasets demonstrate that our approach, which dynamically plans workflows for each query, not only achieves high answer quality but also maintains both cost and latency within acceptable limits.The code of MAO-ARAG is on https://github.com/chenyiqun/Agentic-RAG.

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。

↑