LLM 智能体分布式协调研究：符号引导提升多智能体协作表现
Symbolic Guidance for LLM Agents in Distributed Multiagent Coordination
AI 导读
研究者在 AgentsNet 分布式协调基准上提出符号引导分类法（SGT），将 LLM 智能体的自主性从开放式自然语言推理到完全 prescribed 算法执行划分为多个层级。结果显示，中等自主性水平（部分伪代码引导）稳定优于无引导智能体和完全规定性规范。该工作已作为扩展摘要发表于 AAMAS 2026，指出自主性调节是 LLM 分布式协调的关键设计原则。
正文 · 原文
Abstract:Large language models (LLMs) are increasingly deployed as autonomous agents in multi-agent systems, yet their ability to reliably execute distributed coordination protocols remains poorly understood. While AgentsNet, a benchmark framework for distributed coordination among LLM agents, enables such coordination, granting full reasoning autonomy often leads to inconsistent or degraded performance in complex domains.
We hypothesize that coordination can be improved by regulating agent autonomy through symbolic guidance derived from established algorithms. To investigate this, we introduce the \emph{Symbolic Guidance Taxonomy (SGT)}, which characterizes a spectrum of autonomy ranging from open-ended natural language reasoning to fully prescribed algorithmic execution, with intermediate levels providing partial pseudocode guidance. Our results show that intermediate autonomy levels consistently outperform both unguided agents and fully prescriptive specifications. These findings identify autonomy regulation as a key design principle for LLM-based distributed coordination.
| Comments: | A preliminary version of this work was published as an extended abstract in the Proceedings of AAMAS 2026 |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2609.31963 [cs.AI] |
| (or arXiv:2609.31963v2 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2609.31963 arXiv-issued DOI via DataCite |
Submission history
From: Ben Rachmut [view email]
[v1]
Fri, 25 Sep 2026 20:07:57 UTC (291 KB)
[v2]
Thu, 22 Oct 2026 18:20:41 UTC (293 KB)
来源：arXiv cs.AI · arxiv.org