Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Li, Yiping, An, Zhiyu, Du, Wan
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2604.13349
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918447388033024
author Li, Yiping
An, Zhiyu
Du, Wan
author_facet Li, Yiping
An, Zhiyu
Du, Wan
contents Communication in Large Language Model (LLM)-based multi-agent systems is moving beyond discrete tokens to preserve richer context. Recent work such as LatentMAS enables agents to exchange latent messages through full key-value (KV) caches. However, full KV relay incurs high memory and communication cost. We adapt eviction-style KV compression to this setting and introduce Orthogonal Backfill (OBF) to mitigate information loss from hard eviction. OBF injects a low-rank orthogonal residual from discarded KV states into the retained KV states. We evaluate proposed method against full KV relay on nine standard benchmarks spanning mathematical reasoning, coding, and knowledge-intensive QA. It achieves performance comparable to full KV relay while reducing communication cost by 79.8%--89.4%. OBF further improves the performance and achieves the best results on 7 of the 9 benchmarks. This suggests that more information does not necessarily lead to better communication; preserving the most useful information matters more. Our codebase is publicly available on https://github.com/markli404/When-Less-Latent-Leads-to-Better-Relay.
format Preprint
id arxiv_https___arxiv_org_abs_2604_13349
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle When Less Latent Leads to Better Relay: Information-Preserving Compression for Latent Multi-Agent LLM Collaboration
Li, Yiping
An, Zhiyu
Du, Wan
Machine Learning
Communication in Large Language Model (LLM)-based multi-agent systems is moving beyond discrete tokens to preserve richer context. Recent work such as LatentMAS enables agents to exchange latent messages through full key-value (KV) caches. However, full KV relay incurs high memory and communication cost. We adapt eviction-style KV compression to this setting and introduce Orthogonal Backfill (OBF) to mitigate information loss from hard eviction. OBF injects a low-rank orthogonal residual from discarded KV states into the retained KV states. We evaluate proposed method against full KV relay on nine standard benchmarks spanning mathematical reasoning, coding, and knowledge-intensive QA. It achieves performance comparable to full KV relay while reducing communication cost by 79.8%--89.4%. OBF further improves the performance and achieves the best results on 7 of the 9 benchmarks. This suggests that more information does not necessarily lead to better communication; preserving the most useful information matters more. Our codebase is publicly available on https://github.com/markli404/When-Less-Latent-Leads-to-Better-Relay.
title When Less Latent Leads to Better Relay: Information-Preserving Compression for Latent Multi-Agent LLM Collaboration
topic Machine Learning
url https://arxiv.org/abs/2604.13349