HyLaT: Efficient Multi-Agent Communication via Hybrid Latent-Text Protocol

2026-05-25

AI Summary

This work addresses the trilemma in large language model–based multi-agent systems, where communication efficiency, interpretability, and interaction flexibility are difficult to achieve simultaneously. The authors propose a hybrid latent-variable–text communication protocol that, for the first time, integrates multi-channel communication theory into multi-agent systems. By employing a latent-variable channel to efficiently convey complex cognitive states and supplementing it with natural language to express critical information, the protocol ensures both interpretability and precision. A two-stage training framework—comprising single-agent hybrid generative learning followed by multi-agent interactive co-training—enables significant reductions in communication overhead while maintaining strong performance across diverse task settings. The approach demonstrates notable generalization capability and robustness without sacrificing communicative fidelity.

Abstract

Communication protocol design is a central challenge in large language model-based multi-agent systems. Existing single-channel approaches face an inherent communication trilemma: text-based methods are interpretable but verbose, while latent-space methods are efficient but opaque and limited to unidirectional workflows. Inspired by multi-channel communication theory, we propose HyLaT, a hybrid latent-text communication protocol that transmits elaborate cognitive signals through a latent channel for efficiency, while expressing concise critical signals in natural language to preserve interpretability and precision. We introduce a two-stage training framework combining single-agent hybrid generation learning and multi-agent interactive co-training, enabling agents to generate and interpret hybrid messages across multiple rounds of interaction. Experiments demonstrate that HyLaT reduces communication overhead significantly while maintaining competitive task performance, with strong generalization and robustness across diverse settings.

Problem

• multi-agent communication
• communication protocol
• latent space
• interpretability
• efficiency

Innovation

• hybrid communication
• latent-text protocol
• multi-agent systems
• communication efficiency
• interpretability