Recursive or looped language models have recently emerged as a new scaling axis by iteratively refining the same model computation over latent states to deepen reasoning. We extend such scaling principle from a single model to multi-agent systems, and ask: Can agent collaboration itself be scaled through recursion? To this end, we introduce RecursiveMAS, a recursive multi-agent framework that casts the entire system as a unified latent-space recursive computation. RecursiveMAS connects heterogeneous agents as a collaboration loop through the lightweight RecursiveLink module, enabling in-distribution latent thoughts generation and cross-agent latent state transfer. To optimize our framework, we develop an inner-outer loop learning algorithm for iterative whole-system co-optimization through shared gradient-based credit assignment across recursion rounds. Theoretical analyses of runtime complexity and learning dynamics establish that RecursiveMAS is more efficient than standard text-based MAS and maintains stable gradients during recursive training. Empirically, we instantiate RecursiveMAS under 4 representative agent collaboration patterns and evaluate across 9 benchmarks spanning mathematics, science, medicine, search, and code generation. In comparison with advanced single/multi-agent and recursive computation baselines, RecursiveMAS consistently delivers an average accuracy improvement of 8.3%, together with 1.2$\times$-2.4$\times$ end-to-end inference speedup, and 34.6%-75.6% token usage reduction. Code and Data are provided in https://recursivemas.github.io.

翻译：递归或循环式语言模型最近作为新的扩展维度出现，通过在潜在状态上迭代细化相同模型计算来加深推理。我们将这种扩展原则从单一模型扩展到多智能体系统，并探讨：智能体协作本身能否通过递归实现扩展？为此，我们提出RecursiveMAS，一种递归式多智能体框架，将整个系统视为统一的潜在空间递归计算。RecursiveMAS通过轻量级RecursiveLink模块将异构智能体连接为协作循环，实现分布内潜在思维生成和跨智能体潜在状态传输。为优化我们的框架，我们开发了内外循环学习算法，通过跨递归轮次的共享梯度分配实现迭代式的整体系统协同优化。运行时复杂度与学习动态的理论分析表明，RecursiveMAS比基于文本的标准多智能体系统更高效，且在递归训练中能维持稳定的梯度。实验层面，我们在4种典型智能体协作模式下实例化RecursiveMAS，并在涵盖数学、科学、医学、搜索和代码生成的9个基准上进行评估。与先进单/多智能体及递归计算基线相比，RecursiveMAS持续实现了平均8.3%的准确率提升，同时获得1.2倍至2.4倍的端到端推理加速和34.6%-75.6%的令牌使用量降低。代码与数据见https://recursivemas.github.io。