arXiv：cs.LG· Dongsen Zhang, Peipei Li, Zekun Li, Wenjun Xu·· 4 小时前AI 评分44
KITE：面向任务的关键层 KV 通信实现高效隐式多智能体协作
Task-Oriented Key-Layer KV Communication for Efficient Latent Multi-Agent Collaboration
AI 导读
KITE 是一个免训练框架，将隐式通信目标从发送端状态保真转向接收端任务充分性，通过接收端轨迹失真准则识别任务有效的关键层，只传输该层隐式工作记忆并以其作为自回归隐式推理入口。在两类模型家族、三种模型规模共七个基准上，相比全层 KV 通信，KITE 通信量减少 28-36 倍，端到端推理最高加速 3 倍，准确率最多提升 23.3 个百分点。
正文
Abstract:Large language model-based multi-agent systems improve complex problem solving through collaboration, while latent communication directly transmits model internal states to avoid the high inference costs of natural language. However, existing KV-based latent communication methods prioritize sender-side state fidelity, leading to substantial communication and computation overhead and potentially introducing redundant information. To address these limitations, we revisit latent communication from a task-oriented perspective, shifting its objective from sender-side state fidelity to receiver-side task sufficiency. Under this formulation, we propose KITE, a training-free framework for task-oriented key-layer KV communication. KITE identifies a task-effective key layer using a receiver trajectory distortion criterion, transmits only the latent working memory associated with the key layer, and further uses the same layer as the entry point for autoregressive latent reasoning. Experiments on seven benchmarks across two model families and three model scales show that, compared with full-layer KV communication, KITE reduces communication volume by 28-36$\times$, achieves up to 3$\times$ end-to-end inference speedup, and improves accuracy by up to 23.3 percentage points.
| Subjects: | Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.08820 [cs.LG] |
| (or arXiv:2610.08820v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.08820 arXiv-issued DOI via DataCite |
Submission history
From: DongSen Zhang [view email]
[v1]
Thu, 24 Sep 2026 18:11:27 UTC (5,177 KB)
来源：arXiv：cs.LG · arxiv.org