发表机构
University of Arizona; Hong Kong Baptist University; Amazon Web Services（亚利桑那大学; 香港浸会大学; 亚马逊云服务）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文提出接收者相对有界协调框架，证明贝叶斯充分压缩不足，通过三阶段分解和单交叉条件，指导推理时选择器在准确率与成本间优化通信。
AI 中文摘要
多智能体大语言模型系统将具有广泛上下文的发送者与具有有限局部视图的执行者配对。我们研究了短消息何时能改善执行者的下一个决策，原始上下文何时更可取，以及更强的发送者何时有帮助。我们的框架，\u201c接收者相对有界协调\u201d，将消息效用表示为接收者增益减去协议税。当税收节省超过因信息遗漏和解码器不匹配造成的损失时，压缩优于原始上下文。即使贝叶斯充分压缩，当有界执行者无法使用其表面形式时也可能失败。一个三阶段分解将外部化、吸收和\u201c行动闭合\u201d分开，解释了在正确内容到达接收者后错误如何仍然存在。在单交叉条件下，发送者升级在接收者负担阈值以上有帮助。在六个基准测试中，相同的Qwen协议将ContextBench联合准确率从0.633提高到0.775，但将ToolSandbox从0.889降低到0.653。固定消息重放揭示了尽管工件恢复正确，仍存在闭合失败。这些结果指导了一个推理时选择器，在评估的通信机制上改善了准确率-成本前沿。
英文摘要
Multi-agent LLM systems pair a sender with broad context and an executor with a limited local view. We study when a short message improves the executor's next decision, when raw context is preferable, and when a stronger sender helps. Our framework, \emph{receiver-relative bounded coordination}, expresses message utility as receiver gain minus protocol tax. Compression beats raw context when tax savings exceed losses from omitted information and decoder mismatch. Even \emph{Bayes-sufficient} compression can fail when a bounded executor cannot use its surface form. A three-stage decomposition separates externalization, absorption, and \emph{action closure}, explaining how errors remain after the correct content reaches the receiver. Under a single-crossing condition, sender upgrades help above a receiver-burden threshold. Across six benchmarks, the same Qwen protocol raises ContextBench joint accuracy from $0.633$ to $0.775$ but lowers ToolSandbox from $0.889$ to $0.653$. Fixed-message replay reveals closure failures despite correct artifact recovery. These results guide an inference-time selector that improves the accuracy-cost frontier on the evaluated communication regimes.
CommentsPublished at Neurips 2026
