发表机构
Texas A&M University; Brookhaven National Laboratory（德克萨斯A&M大学; 布鲁克海文国家实验室）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究从信息瓶颈视角探讨多智能体系统（MAS）与单智能体系统（SAS）差异，指出MAS优势在有限中继时因压缩权衡产生，由参数β控制，通过实验验证，解释了有限智能体间通信何时有利有害。
AI 中文摘要
由大语言模型驱动的多智能体系统（MAS）已成为处理复杂任务的一种有前景的范式。但其相对于单智能体系统（SAS）的优势仍不明确，性能在不同设置下变化不一致。本文从信息瓶颈角度阐释MAS与SAS的差异。关键观察是SAS在一个共享上下文中累积完整推理轨迹，而MAS使用由有限中继消息连接的孤立局部上下文。研究表明，在无限中继带宽下，任何SAS都可由传输完整上游上下文的MAS模拟。MAS的显著优势出现在有限中继情况下，此时压缩带来基本权衡：减少冗余上下文可提高效率，但可能导致任务相关信息丢失。将此权衡形式化为由有效参数β控制的信息瓶颈，β体现平衡如何随模型能力变化。当上下文减少超过中继信息丢失时，MAS有优势。通过五个基准和三个模型规模进行18个对照实验验证理论研究。观察到当中继接近充足时，MAS始终有帮助，尤其是对较弱模型；当中继导致信息丢失时，MAS优势缩小或反转，尤其是对能从冗余上下文中提取有用信息的较强模型。研究表明多智能体设计本质上是信息瓶颈优化问题，该视角解释了有限智能体间通信何时有帮助或有损害。
英文摘要
LLM powered multi-agent systems (MAS) have emerged as a promising paradigm for complex tasks. However, their advantages over single-agent systems (SAS) remain unclear, with performance varying inconsistently across settings. Here, we provide an information bottleneck perspective on elucidating the differences between MAS and SAS. Specifically, our key observation is that a SAS accumulates its full reasoning trace in one shared context, while a MAS uses isolated local contexts connected by bounded relay messages. We show that, under infinite relay bandwidth, any SAS can be simulated by a MAS that transmits the full upstream context. Thus, the nontrivial advantage of MAS arises under bounded relays, where compression introduces a fundamental trade-off: reducing redundant context can improve efficiency, but may also incur loss of task-relevant information. We formalize this trade-off as an information bottleneck controlled by an effective parameter $β$, which captures how the balance shifts with model capability, and shows that MAS gains arise when context reduction outweighs relay information loss. We conduct 18 controlled experiments across five benchmarks and three model scales to validate our theoretical studies. We observe that MAS consistently helps when relays are near-sufficient, especially for weaker models. In contrast, MAS gains shrink or reverse when relays incur information loss, especially for stronger models that can already extract useful information from redundant context and thus gain little from compression. Our study shows that multi-agent design is fundamentally an information-bottleneck optimization problem. This perspective explains when bounded inter-agent communication helps or hurts.
