mlsanityData to 7 Oct 2026

Faithful, Not Corrective: Model Capability Governs Message-Format Effects in Multi-Hop Agent Relays

arXiv 2607.09678 ↗First posted 12 Jun 2026cs.AI cs.LGcited by 0

On the map: LLM agents and tool use → General LLM agent frameworks and benchmarks

Abstract (arXiv)

When LLM agents hand information to one another, does the message format matter? Two literatures disagree: format-optimization work reports that structured messages cut cost without hurting accuracy, while format-restriction studies find that imposing structure degrades generation. Neither line has measured what happens when messages traverse multiple hops, where copy fidelity, rather than one-shot generation quality, dominates. We introduce a controlled relay testbed in which briefs of twelve programmatic atomic facts are re-encoded hop by hop in five formats (free natural language, precision-instructed NL, JSON, triples, key-value) over six hops, scored against programmatic ground truth by a fixed strong grader, across two relay-capability tiers, a cognitive-load condition, and a paired-fork error injection. We find that (i) a strong relay is nearly lossless for every format (hop-6 QA recall ≥ 0.973), with residual loss concentrated at the first encoding step; (ii) per-hop cognitive load raises generation cost by 24-53% while fidelity changes stay within ± 1.8 points; (iii) under a weak 1.5B relay, the across-format dispersion of hop-6 recall grows by a factor of 8.7 (CI 5.3-15.5), driven by an encode-drift trade-off that flips the format ranking in transit; and (iv) once an injected error is present, every format propagates it faithfully (surface persistence 83-100%) and no format cascades collateral damage onto neighboring facts. Structure buys a faithful, error-localizing channel, not an error-correcting code.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in General LLM agent frameworks and benchmarks, one of 16 directions in LLM agents and tool use. Builds on 12 of its 28 references that are in the map. First posted on arXiv 12 Jun 2026.

Builds on

12 of 28 references.

Closest ancestors · 3

Same direction, newest first

  1. 2026Single-Agent LLMs Outperform Multi-Agent Systems on Multi-Hop Reasoning Under Equal Thinking Token Budgets2 Apr
  2. 2026Information Fidelity in Tool-Using LLM Agents: A Martingale Analysis of the Model Context Protocol10 Feb
  3. 2025Towards a Science of Scaling Agent Systems9 Dec

Other references · 9

  1. 2023Improving Factuality and Reasoning in Language Models through Multiagent Debatecited by 1,739
  2. 2025A survey of agent interoperability protocols: Model Context Protocol (MCP), Agent Communication Protocol (ACP), Agent-to-Agent Protocol (A2A), and Agent Network Protocol (ANP)cited by 98
  3. 2025KVComm: Enabling Efficient LLM Communication through Selective KV Sharingcited by 16
  4. 2024Beyond Natural Language: LLMs Leveraging Alternative Formats for Enhanced Reasoning and Communicationcited by 14
  5. 2025LLM as a Broken Telephone: Iterative Generation Distorts Informationcited by 12
  6. 2025The PIMMUR Principles: Ensuring Validity in Collective Behavior of LLM Societiescited by 12
  7. 2025Toward a Safe Internet of Agentscited by 6
  8. 2025Natural Language Tools: A Natural Language Approach to Tool Calling In Large Language Agentscited by 3
  9. 2026Lost Before Translation: Social Information Transmission and Survival in AI-AI Communicationcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026When Less Latent Leads to Better Relay: Information-Preserving Compression for Latent Multi-Agent LLM Collaboration
  2. 2026When Do Multi-Agent Systems Help? An Information Bottleneck Perspective
  3. 2026Do Latent Channels Actually Communicate? A Causal Audit of Latent Multi-Agent LLM
  4. 2026When Does Latent Communication Pay? A Causal Audit of Relayed KV Caches in Multi-Agent LLMs
  5. 2026Bayes-Sufficient Compression Is Not Enough: How Does Communication Help Multi-Agent Systems?
  6. 2026Harness or Model? Isolating the Harness Effect in Agentic Coding with a Contamination-Controlled Private Suite
  7. 2025Benefits and Limitations of Communication in Multi-Agent Reasoning
  8. 2025AsymPuzl: An Asymmetric Puzzle for multi-agent cooperation