让多个智能体的协作网络在运行中自动进化，提升复杂任务解决能力。
MANTA: Multi-Agent Network Topology Adaptation for Self-Evolving Multi-Agent Systems
- 基于任务动态调整智能体间的通信结构，支持角色、链接、顺序等实时变化。
- 在5个基准上平均得分74.0，领先最强基线5.8个百分点，PlanCraft表现最佳。
- 适合需要自适应协作的多智能体系统研究者，尤其关注推理时结构演化。
基于大语言模型的多智能体系统通过任务分解、角色分化、信息交换和中间验证来提升复杂问题求解能力。然而，现有系统通常将通信拓扑视为固定设计或离线优化目标。本文提出MANTA框架，实现多智能体网络拓扑的自适应演进，使通信结构能在推理时自我更新。部署前，MANTA从先前结构经验中初始化任务相关的拓扑；运行中，监控协作轨迹，在当前组织失效时进行有限度的结构更新。这些更新可调整智能体角色、通信链路、执行顺序、信息可见性及验证路径，同时保持任务接口与智能体预算不变。我们在涵盖信息检索、工具使用、规划、工作流执行和数学推理的五个基准上评估MANTA，结果表明其平均得分74.0，优于最强基线5.8个百分点，并在PlanCraft上取得最优成绩。结果证明，推理时的自优化可扩展至协作架构本身。
原文摘要 · Abstract (English)
Large language model-based multi-agent systems improve complex problem solving through task decomposition, agent specialization, information exchange, and intermediate validation. However, existing systems typically treat communication topology as a fixed design choice or an offline optimization target. We introduce MANTA, a framework for Multi-Agent Network Topology Adaptation that enables communication structures to self-evolve at inference time. Before execution, MANTA initializes a task-conditioned topology from prior structural experience. During deployment, it monitors collaboration traces and applies bounded structural updates when the current organization becomes insufficient. These updates can modify agent roles, communication links, execution order, information visibility, and validation pathways while preserving the task interface and agent budget. We evaluate MANTA against representative single-agent and multi-agent baselines on five benchmarks spanning information seeking, tool use, planning, workflow execution, and mathematical reasoning. MANTA achieves the highest average score of 74.0, outperforming the strongest baseline by 5.8 percentage points and obtaining the best result on PlanCraft. These results show that inference-time self-improvement can extend to the architecture of collaboration itself.
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。