{
  "abstract": "Multi-agent LLM systems pair a sender with broad context and an executor with a limited local view. We study when a short message improves the executor's next decision, when raw context is preferable, and when a stronger sender helps. Our framework, \\emph{receiver-relative bounded coordination}, expresses message utility as receiver gain minus protocol tax. Compression beats raw context when tax savings exceed losses from omitted information and decoder mismatch. Even \\emph{Bayes-sufficient} compression can fail when a bounded executor cannot use its surface form. A three-stage decomposition separates externalization, absorption, and \\emph{action closure}, explaining how errors remain after the correct content reaches the receiver. Under a single-crossing condition, sender upgrades help above a receiver-burden threshold. Across six benchmarks, the same Qwen protocol raises ContextBench joint accuracy from $0.633$ to $0.775$ but lowers ToolSandbox from $0.889$ to $0.653$. Fixed-message replay reveals closure failures despite correct artifact recovery. These results guide an inference-time selector that improves the accuracy-cost frontier on the evaluated communication regimes.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2610.03769",
  "author_total": 6,
  "authors": [
    {
      "name": "Yi Xie"
    },
    {
      "name": "Zhanke Zhou"
    },
    {
      "name": "Yi Fan"
    },
    {
      "name": "Yong Ge"
    },
    {
      "name": "Bo Han"
    },
    {
      "name": "Bo Liu"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/8e52fwygwa3dn.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2610.03769",
  "fulltext_url": null,
  "id": "8e52fwygwa3dn",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2610.03769",
  "language": null,
  "licence": "CC-BY-4.0",
  "licence_label": "CC BY 4.0",
  "licence_source": "article",
  "licence_url": "http://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": "CC0-1.0",
  "native_subjects": [
    [
      "arxiv",
      "cs.IT"
    ],
    [
      "arxiv",
      "cs.LG"
    ],
    [
      "arxiv",
      "cs.MA"
    ],
    [
      "arxiv",
      "math.IT"
    ],
    [
      "openalex",
      "1702|Multi-Agent Systems and Negotiation"
    ]
  ],
  "openalex": "W7220723244",
  "pdf_url": "https://arxiv.org/pdf/2610.03769",
  "pmcid": null,
  "pmid": null,
  "published": "2026-10-01",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/8e52fwygwa3dn.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:27:10Z",
      "native_id": "2610.03769",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2610.03769"
    }
  ],
  "subjects": [
    {
      "field": "mathematics",
      "origin": "arxiv:math.IT",
      "subfield": "general"
    }
  ],
  "title": "Bayes-Sufficient Compression Is Not Enough: How Does Communication Help Multi-Agent Systems?",
  "title_text": "Bayes-Sufficient Compression Is Not Enough: How Does Communication Help Multi-Agent Systems?",
  "type": "preprint",
  "updated_at": "2026-10-30T12:27:10Z",
  "url": "https://research.draconitic.com/paper/8e52fwygwa3dn",
  "venue": "arXiv",
  "versions": []
}