用多个小智能体分离思考与工具操作，降低复杂任务中的认知负担。
Reducing Cognitive Overhead in Tool Use via Multi-Small-Agent Reinforcement Learning
- 拆分思考与工具使用：一个推理代理规划，多个工具代理专注执行。
- 数学求解任务中准确率提升，比单大模型方法更稳定。
- 适合需要高可靠性的多工具协作场景，如自动化科研或编程助手。
近期多智能体系统进展表明，通过分工协作的专用小智能体具有潜力。然而，现有工具集成推理系统多采用单智能体范式，由单一大模型交替进行长序列推理与精确工具操作，导致认知负荷干扰和协调不稳定。本文提出MSARL框架——一种多小智能体强化学习方法，显式分离推理与工具使用。在该框架中，推理代理负责问题分解与工具调用规划，多个工具代理则分别专注于特定外部工具，均通过模仿学习与强化学习结合训练，并使用角色专属奖励。在支持代码执行的数学问题求解任务中，MSARL显著提升了推理稳定性与最终答案准确率，优于单智能体基线。此外，该架构可泛化至多种工具使用任务，证明了小智能体的认知角色解耦是可扩展的多智能体人工智能设计蓝图。
原文摘要 · Abstract (English)
Recent advances in multi-agent systems highlight the potential of specialized small agents that collaborate via division of labor. Existing tool-integrated reasoning systems, however, often follow a single-agent paradigm in which one large model interleaves long-horizon reasoning with precise tool operations, leading to cognitive-load interference and unstable coordination. We present MSARL, a Multi-Small-Agent Reinforcement Learning framework that explicitly decouples reasoning from tool use. In MSARL, a Reasoning Agent decomposes problems and plans tool invocations, while multiple Tool Agents specialize in specific external tools, each trained via a combination of imitation learning and reinforcement learning with role-specific rewards. On mathematical problem solving with code execution, MSARL significantly improves reasoning stability and final-answer accuracy over single-agent baselines. Moreover, the architecture generalizes to diverse tool-use tasks, demonstrating that cognitive-role decoupling with small agents is a scalable blueprint for multi-agent AI design.
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。