让智能体自主协作，解决复杂问题更准更快。
DeAR: Decentralized Agentic Reasoning via Capability Grounding and Collaborative Thought Navigation
- 智能体通过自适应分工，按任务需求动态调整角色。
- 在9个基准测试中表现优于现有方法，准确率显著提升。
- 适合需要多智能体协同推理的复杂知识任务场景。
现有智能体推理系统多依赖中心化协议，导致路由瓶颈和角色固定，难以应对复杂多模态任务。本文提出DeAR（去中心化智能体推理）框架，实现智能体间的自主对等协作。其核心机制包括：（1）基于查询的去中心化能力定位，实现任务驱动的智能体专业化；（2）思想图导航，精准引导智能体间交互；（3）拓扑更新，动态修正错误。在9个涵盖多模态推理与文本问答的基准上评估表明，DeAR持续超越近期基线方法，验证了去中心化、自适应协作在知识密集型推理任务中的有效性。代码将于录用后公开。
原文摘要 · Abstract (English)
Existing agentic reasoning systems typically rely on centralized protocols. This design introduces routing bottlenecks and static role allocations that often fail when handling complex multimodal queries. We propose DeAR (Decentralized Agentic Reasoning), a framework that shifts from central control to autonomous peer-to-peer collaboration. DeAR is built on three mechanisms: (1) decentralized capability grounding for query-dependent agent specialization, (2) thought map navigation for targeted peer interactions, and (3) topology update for adaptive error correction. Evaluations across 9 diverse multimodal reasoning and text-based QA benchmarks indicate that DeAR consistently outperforms recent baseline methods, validating that decentralized and adaptive collaboration among agents enhances accuracy in knowledge-intensive reasoning tasks. The source code will be available at https://open_upon_acceptance.
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。