Metraipolis › archive › dc_ce7b6790cf81

Learning Calibrated Confidence for Multi-Agent Collaboration — arXiv:2610.39927

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.39927
License (as archived)
CC0-1.0
Archived
2026-10-31T06:06:50.642Z
SHA-256 of the archived text
937dfd0bb80f18aff68670492712b2555c5bb4100b624a6d154f3b534f6d0fa3
Stable URL of this page
https://metraipolis.com/documento/dc_ce7b6790cf81

How to cite

Learning Calibrated Confidence for Multi-Agent Collaboration — arXiv:2610.39927 — arXiv (en), https://arxiv.org/abs/2610.39927 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ce7b6790cf81 (Metraipolis archive dc_ce7b6790cf81, sha256 937dfd0bb80f18aff68670492712b2555c5bb4100b624a6d154f3b534f6d0fa3)
BibTeX
@misc{dcce7b6790cf81,
  title = {Learning Calibrated Confidence for Multi-Agent Collaboration — arXiv:2610.39927},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.39927},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ce7b6790cf81; sha256 937dfd0bb80f18aff68670492712b2555c5bb4100b624a6d154f3b534f6d0fa3},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.39927. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Complex information seeking on the Web often requires collecting evidence from multiple documents and reasoning across them, as in multi-hop question answering. Multi-agent systems address such tasks by dividing work among specialized agents, where upstream agents produce intermediate answers and supporting evidence that downstream agents build upon. However, upstream answers vary in correctness, and most systems provide no quantitative signal about it, leaving the downstream agent to judge whether to trust or verify what it receives based solely on the textual evidence. In this paper, we propose MASCAL. Each agent learns to produce a correctness probability for its answer by jointly optimizing task performance and confidence calibration in reinforcement learning; the upstream agent passes this probability at handoff, giving the downstream agent a quantitative basis for its decisions, and the final answer also carries a calibrated probability, quantifying the reliability of the system output. To guide the downstream agent in using this signal during training, we introduce confidence-aware weighting, which increases the learning weight at positions responsive to the upstream probability, so that the agent learns more from decisions where it actively uses the signal. Across three collaboration architectures spanning Web retrieval and reasoning tasks, MASCAL improves task accuracy and calibration quality for both agents, and generalizes to deeper collaboration chains and different model scales. Intervention analysis confirms that upstream confidence affects downstream decision behavior, and the learned final confidence supports selective answering and targeted rechecking.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.MA, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA).
Pagina del lavoro: https://arxiv.org/abs/2610.39927 — 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 2026-10-31T06:06:38.433Z dalla pagina https://arxiv.org/list/cs.MA/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).