Metraipolis › archive › dc_a24a749ff0ca

Beyond Final Accuracy: Auditing Communication in LLM Multi-Agent Systems — arXiv:2610.01042

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.01042
License (as archived)
CC0-1.0
Archived
2026-10-06T12:03:34.259Z
SHA-256 of the archived text
cee7a747d1c63ee56ca73ec8088164734b3a2b3c97ce31bf0cbdd4c08b85f073
Stable URL of this page
https://metraipolis.com/documento/dc_a24a749ff0ca

How to cite

Beyond Final Accuracy: Auditing Communication in LLM Multi-Agent Systems — arXiv:2610.01042 — arXiv (en), https://arxiv.org/abs/2610.01042 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a24a749ff0ca (Metraipolis archive dc_a24a749ff0ca, sha256 cee7a747d1c63ee56ca73ec8088164734b3a2b3c97ce31bf0cbdd4c08b85f073)
BibTeX
@misc{dca24a749ff0ca,
  title = {Beyond Final Accuracy: Auditing Communication in LLM Multi-Agent Systems — arXiv:2610.01042},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.01042},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a24a749ff0ca; sha256 cee7a747d1c63ee56ca73ec8088164734b3a2b3c97ce31bf0cbdd4c08b85f073},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.01042. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Multi-agent communication aims to help agents benefit from one another's information. Yet improvements in system performance leave a fundamental ambiguity: do they reflect effective communication, a favorable agent architecture, or simply additional reasoning? Because communication methods are commonly evaluated within the systems they were designed for, these factors are difficult to disentangle. Final accuracy further merges corrected errors and corrupted answers into a single outcome, obscuring how communication changes decisions. We introduce Independent--Communicate--Revise (ICR), a controlled framework that evaluates communication as answer revision following independent reasoning. ICR fixes initial reasoning trajectories, measures correction and preservation conditional on both agents' initial correctness, and uses a no-message revision control to quantify gains beyond additional reasoning. Across four reasoning benchmarks, our audit of textual and latent communication reveals that similar aggregate accuracy can conceal substantially different revision behaviors. Compared with transmitting answers alone, full reasoning increases correction while reducing preservation on all four benchmarks, so richer messages amplify beneficial and harmful influence alike. Receiver-policy comparisons on MedQA and GPQA-D further show that a structured verification policy shifts every channel toward greater preservation and lower correction, while its effect on selectivity varies across channels and tasks. These findings challenge treating communication quality as an intrinsic property of a channel. ICR therefore recenters evaluation on selective revision, providing a unified framework for examining how message content and receiver policies jointly produce benefits and harms.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-02 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2610.01042 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-06T10:21:27.907Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-10-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).