Preprint

StateBridge: Training-free Hidden-state Alignment for Latent Communication in LLM Multi-Agent Systems

Yan-Wen Peng Delvin Ce Zhang Xi Wang Nikolaos Aletras

Aug 2026 · 6 citations · 41 references

Computer Science

TL;DR

This work proposes 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 and prepended to the input of the receiver agent as a continuous prefix.

Abstract

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.

arXiv Semantic Scholar

View source

Similar papers

Jul 2026

TransMem: Transforming Hidden States into Memory for Large Language Models

TransMem is proposed, a lightweight inference-time parametric memory module that transforms sparse historical hidden states from a frozen LLM backbone into reusable memory representations and introduces evidence-conditioned self-distillation to learn transferable memory utilization rather than task-specific knowledge.

Haodong Lei, Junming Liu, Yirong Chen et al. · arXiv.org · 0 citations

Preprint Aug 2026

Post-Hoc Sparse Coding of Latent Communication Between Vision-Language Model Agents

This work studies Vision Wormhole's redundancy by fitting a post-hoc sparse autoencoder to frozen Vision Wormhole activations and measuring reconstruction, downstream utility, feature reuse, and token-level interventions across nine reasoning benchmarks, and establishes strong post-hoc compressibility relative to the o...

Di Wu, Xiao-Hui Zhu · 1 citation

#artificial intelligence Preprint Aug 2026

Listen to the Latents: Self-Correcting Speech Recognition in Large Audio Language Models Through Hidden-State Interactions

This work refine warm-initialized LLM-based ASR models by leveraging their own pre-adaptation base LLMs by leveraging their own pre-adaptation base LLMs, focusing on LoRA-adapted settings where the base LLM is preserved.

Chan-Jan Hsu, Jaeyeon Kim, Chao-Han Huck Yang et al. · 0 citations

#artificial intelligence Preprint Aug 2026

HSRM: Hidden-State Reward Models for Test-Time Verification

HSRM is introduced, a lightweight hidden-state reward model that verifies candidate solutions by directly reading the generator's internal representations rather than re-processing its text, providing an efficient alternative to text-only verification by reusing representations already computed during generation.

Xianzhi Li, Xiao-Dan Zhu · 0 citations

Preprint Aug 2026

Architecture-Dependent Causal Transfer of Activation States Across Large Language Models

End-to-end activation-state transfer between LLMs, as currently implemented, is architecture-dependent rather than universal, and it is concluded that end-to-end activation-state transfer between LLMs is architecture-dependent rather than universal.

Fernando Cardenas Piepereit · 3 citations

#machine learning Preprint Sep 2026

Large Language Models Develop Belief State Geometry In-Context

Large language models trained on next-token prediction exhibit remarkable in-context learning (ICL) abilities, yet the representations that support ICL remain poorly understood, and representation-level evidence that ICL in open-source LLMs approximates optimal Bayesian prediction over a context-inferred generative mod...

Daniel Balcells, Andrew Lee, Chirag Rastogi et al. · 1 citation · ⚡1