无需训练的隐状态对齐，让大模型代理直接传递连续信息
StateBridge: Training-free Hidden-state Alignment for Latent Communication in LLM Multi-Agent Systems
- 用闭合形式正交变换对齐发送方与接收方的隐藏状态
- 在26个任务中22个表现最优，超越现有最强基线
- 无需额外训练，兼容预训练模型输入分布，适合快速部署
基于大语言模型的多智能体系统通常通过文本（离散标记）进行通信，但文本引入了离散瓶颈。将发送方的连续隐藏状态转换为离散标记会丢失仅靠标记身份无法捕捉的信息。近期工作提出潜在通信作为替代方案，即智能体直接传输隐藏表示而不转为文本。然而，现有方法要么逐层注入工作记忆模块，要么需要训练投影器，限制了可移植性。本文提出StateBridge，一种无需训练的潜在通信方法，通过闭合形式正交变换将发送方最终层隐藏状态对齐至接收方输入空间。轻量级归一化校准和词汇锚定确保与预训练输入分布兼容。对齐后的状态以连续前缀形式拼接到接收方输入中。我们在两个模型族的四个模型上评估了Math Reasoning、Code Generation和Question Answering任务。StateBridge在26个模型-任务组合中，有22个达到最佳或并列最佳成绩，持续优于最强基线。
原文摘要 · Abstract (English)
Large language model based multi-agent systems usually communicate in text, i.e., using discrete tokens. However, text introduces a discrete bottleneck. Converting the sender's continuous hidden states into discrete tokens discards information that token identities alone cannot capture. Recent work proposes latent communication as an alternative, where agents transmit hidden representations directly without converting them to text. However, existing latent methods either inject working memory layer by layer across the transformers, or require trained projectors that limit portability. We propose StateBridge, a training-free latent communication approach that aligns the sender's final-layer hidden states to the receiver's input space via a closed-form orthogonal transformation. Lightweight norm calibration and vocabulary anchoring ensure compatibility with the pretrained input distribution. The aligned states are prepended to the input of the receiver agent as a continuous prefix. We evaluate StateBridge on math reasoning, code generation, and question answering with four models from two families. StateBridge achieves the best or tied-best score on 22 out of 26 model-task pairs, consistently outperforming the strongest baseline.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。