多智能体协同推理，减少视觉任务中的幻觉错误。
Visual Para-Thinker++: A Single-Policy Multi-Agent Framework for Visual Reasoning
- 共享大模型策略，分主、工、总三角色并行推理。
- 在多个基准上优于单路径和并行基线，幻觉敏感任务提升显著。
- 适合需要高可靠视觉推理的场景，如医疗图像分析。
视觉推理需整合分布在区域、属性和关系中的证据，单链推理易过早感知并产生幻觉。我们提出 Visual Para-Thinker++，一种基于共享多模态大模型策略的多智能体框架，其中同一策略被实例化为角色相关的主代理、工作代理和总结代理。主代理以固定模式分解任务；工作代理在上下文隔离下并行推理；总结代理融合完整推理轨迹，而非对最终标签进行多数投票。共享策略通过多智能体能力注入与角色解耦优化训练，为不同角色分配专属奖励与优势，降低协作中的梯度冲突。原生推理引擎通过共享视觉前缀和键值缓存复用实现高效多智能体推演。在 V*、CountBench、RefCOCO 系列及 HallusionBench 上，Visual Para-Thinker++ 持续超越单轨迹与推理时并行基线，尤其在幻觉敏感的视觉推理任务中表现突出。
原文摘要 · Abstract (English)
Visual reasoning requires integrating evidence distributed across regions, attributes, and relations, making single-chain reasoning prone to early perceptual commitment and hallucination. We propose Visual Para-Thinker++, a single-policy multi-agent framework in which one shared MLLM policy is instantiated as role-conditioned Main, Worker, and Summary Agents. The Main Agent decomposes the task with fixed allocation patterns; Worker Agents reason in parallel under context isolation; and the Summary Agent reconciles full Worker reasoning traces rather than majority-voting on final labels. The shared policy is trained by Multi-Agent Capability Injection and Role-Decoupled Multi-Agent Optimization, which assign role-specific rewards and advantages to corresponding token segments to reduce gradient conflict among collaborative roles. A native inference engine enables efficient multi-agent rollout through shared visual prefix and KV cache reuse. Across V*, CountBench, the RefCOCO family, and HallusionBench, Visual Para-Thinker++ consistently outperforms single-trajectory and inference-time parallel baselines, with especially strong gains on hallucination-sensitive visual reasoning.
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。