Multi-agent language systems are often built as hand-designed workflows, where agents are assigned semantic roles and communication protocols are specified in advance. We propose NeuroMAS, a method that first treats a multi-agent language system as a trainable and scalable neural-network-like architecture with LLM agents as nodes and intermediate textual signals as edges. In NeuroMAS, agent nodes are role-free but structure-aware: the topology only determines how information can flow in general, while reinforcement learning training determines how nodes communicate, specialize, and coordinate. This formulation shifts multi-agent design from workflow engineering toward architecture design, where depth, width, connectivity, and growth protocol become scalable sources of capability. Further, we provide a theoretical perspective showing why such modular textual computation is more parameter-efficient when tasks admit hierarchical decompositions. Experiments show that NeuroMAS improves significantly over both inference-time and trained multi-agent baselines. We further find that organizational scaling is path-dependent: larger systems can be challenging to train from scratch, but become feasible when grown progressively from smaller trained systems. These results suggest that learned neural multi-agent systems are a promising scaling axis for LLMs.

翻译：多智能体语言系统通常被设计为手工构建的工作流程，其中智能体被赋予语义角色，通信协议被预先指定。我们提出NeuroMAS方法，该方法首先将多智能体语言系统视为一种可训练且可扩展的类神经网络架构，其中LLM智能体作为节点，中间文本信号作为边。在NeuroMAS中，智能体节点是无角色但结构感知的：拓扑仅决定信息在一般情况下如何流动，而强化学习训练则决定节点如何通信、特化与协调。这种表述将多智能体设计从工作流程工程转向架构设计，其中深度、宽度、连通性和增长协议成为可扩展的能力来源。此外，我们提供一个理论视角，证明当任务允许层次化分解时，这种模块化文本计算在参数效率上更高。实验表明，NeuroMAS相较于推理时和训练后的多智能体基线均有显著提升。我们进一步发现，组织扩展具有路径依赖性：较大的系统从头开始训练可能具有挑战性，但当从较小的训练系统逐步增长时则变得可行。这些结果表明，学习型的神经多智能体系统是LLM一个富有前景的扩展维度。