发表机构
IIIS, Tsinghua University; School of Artificial Intelligence, Shanghai Jiao Tong University; Shanghai Qi Zhi Institute（清华大学交叉信息研究院; 上海交通大学人工智能学院; 上海期智研究院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
CERA-MoA提出协同进化路由与持续学习智能体的迭代强化学习框架，通过预测性熟悉度估计和自适应路由机制，在性能与效率间取得平衡，并优于现有基线。
AI 中文摘要
当前的混合智能体（Mixture-of-Agents, MoA）范式通常将查询路由和智能体微调视为独立过程，限制了其应对不断演化的智能体能力的能力。这种脱节使得路由策略无法在训练后适应智能体能力的演化，也阻碍了智能体实现协同的数据驱动专业化。为解决这一问题，我们提出了CERA-MoA（用于混合智能体的协同进化路由器与持续学习智能体），一种迭代强化学习框架，其中动态路由器与独立智能体策略协同进化。我们设计了一个预测性熟悉度估计器，利用中间层隐藏状态来评估智能体间的语义能力，避免了完整rollout的开销。基于这些熟悉度分数，一种累积阈值自适应路由机制动态激活量身定制的最小智能体子集，在任务性能与效率之间取得平衡。通过根据智能体不断演化的能力主动为其分配针对性的训练样本，CERA-MoA促进了能力差异化。跨多个领域的广泛实验表明，CERA-MoA优于最先进的静态智能体路由和固定工作流微调基线。
英文摘要
Current Mixture-of-Agents (MoA) paradigms generally treat query routing and agent fine-tuning as separate processes, limiting their ability to respond to evolving agent capabilities. This disconnect prevents routing strategies from adapting to evolving agent capabilities during post-training and prevents agents from achieving synergistic data-driven specialization. To resolve this, we introduce CERA-MoA (Co-Evolving Router with continually learning Agents for Mixture-of-Agents), an iterative reinforcement learning framework where the dynamic router and independent agent policies co-evolve. We design a predictive familiarity estimator that leverages mid-layer hidden states to evaluate semantic competence among agents, avoiding the overhead of full rollouts. Based on these familiarity scores, a cumulative-threshold adaptive routing mechanism dynamically activates a tailored minimal agent subset, achieving a trade-off between task performance and efficiency. By proactively allocating targeted training samples to agents based on their evolving competence, CERA-MoA promotes capability differentiation. Extensive experiments across various domains demonstrate that CERA-MoA outperforms state-of-the-art static-agent routing and fix-workflow fine-tuning baselines.
