← 研踪 PaperTrailarXiv:2605.16757 · v3
NeuroMAS: Multi-Agent Systems as Neural Networks with Joint Reinforcement Learning
首次提交 2026-05-16 10:11
版本更新 2026-11-06 06:41
雷达观测 2026-11-09 10:16
摘要
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 framework that treats a multi-agent language system as a trainable and scalable neural-network-like architecture with hidden textual intermediate signals. In NeuroMAS, language-agent nodes are role-free but structure-aware: the topology determines how information can flow, while reward-based 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. We provide a theoretical perspective showing why such modular text-valued computation can be parameter-efficient when tasks admit stable decompositions. Experiments show that NeuroMAS improves over both inference-time collaboration methods and trained baselines, and that its gains are not explained by trainable parameter count alone. We further find that organizational scaling is path-dependent: larger systems can fail when trained from scratch, but become stronger when grown progressively from smaller trained systems. These results suggest that learned multi-agent organization is a promising scaling axis for language-model reasoning and generation.
期刊信息：The First Conference on Statistics and Trustworthy AI for Cross (X)-Domain Acceleration (STAI-X 2026)
观测到的版本
相邻记录可能跳过中间版本；首次发现 v2 及以上也会标为版本更新。
v3 · 2026-11-06 06:41 · 最新观测版本
NeuroMAS: Multi-Agent Systems as Neural Networks with Joint Reinforcement Learning
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 framework that treats a multi-agent language system as a trainable and scalable neural-network-like architecture with hidden textual intermediate signals. In NeuroMAS, language-agent nodes are role-free but structure-aware: the topology determines how information can flow, while reward-based 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. We provide a theoretical perspective showing why such modular text-valued computation can be parameter-efficient when tasks admit stable decompositions. Experiments show that NeuroMAS improves over both inference-time collaboration methods and trained baselines, and that its gains are not explained by trainable parameter count alone. We further find that organizational scaling is path-dependent: larger systems can fail when trained from scratch, but become stronger when grown progressively from smaller trained systems. These results suggest that learned multi-agent organization is a promising scaling axis for language-model reasoning and generation.
查看此版本解读 →查看此版本原文 ↗