BCMAAC：基于信念条件化的多智能体 Actor-Critic 方法实现去中心化 LLM 协作
Belief-Conditioned Multi-Agent Actor Critic for Decentralized LLM Collaboration
AI 导读
一项 arXiv 论文提出 BCMAAC，一种信念条件化的多智能体 Actor-Critic 方法，通过维护对上下文信息的信念来促进去中心化 LLM 协作的训练。
正文 · 原文
Abstract:Recent work has explored optimizing LLM collaboration through Multi-Agent Reinforcement Learning (MARL), but many approaches still rely on predefined workflows, limiting efficiency and scalability. Decentralized LLM collaboration is challenging because agents execute independently with limited communication, share a joint objective, and must infer useful information from communication traces. In this paper, we propose BCMAAC, a Belief-Conditioned Multi-Agent Actor-Critic method that maintains beliefs over contextual information and uses them to facilitate training. Experiments on Minecraft game-playing, web searching and tool-using tasks show that BCMAAC improves task performance and collaborative behavior compared to naive actor-critic baselines. The improvements are particularly pronounced in multi-turn tasks with richer contextual information, where belief modeling provides a clear advantage.
| Comments: | Accepted at the NeurIPS 2026 Workshop PTA: From Pretrained Representations to Acting Agents. 18 pages, 3 figures |
| Subjects: | Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Multiagent Systems (cs.MA) |
| Cite as: | arXiv:2610.35398 [cs.AI] |
| (or arXiv:2610.35398v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.35398 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Tianle Chen [view email]
[v1]
Mon, 26 Oct 2026 13:12:41 UTC (1,042 KB)
来源：arXiv cs.AI · arxiv.org