Metraipolis › archive › dc_f13c4495cd4b
Online Verification of Language Model Responses Under Cost Constraints — arXiv:2610.02632
- Source
- arXiv
- Original
- https://arxiv.org/abs/2610.02632
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-05T23:09:43.360Z
- SHA-256 of the archived text
8b3bf5e31b07afee2b3795c0fc9318c68c5c073eb7d67112ec94cf797b1d9414- Stable URL of this page
https://metraipolis.com/documento/dc_f13c4495cd4b
How to cite
Online Verification of Language Model Responses Under Cost Constraints — arXiv:2610.02632 — arXiv, https://arxiv.org/abs/2610.02632 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_f13c4495cd4b (Metraipolis archive dc_f13c4495cd4b, sha256 8b3bf5e31b07afee2b3795c0fc9318c68c5c073eb7d67112ec94cf797b1d9414)
BibTeX
@misc{dcf13c4495cd4b,
title = {Online Verification of Language Model Responses Under Cost Constraints — arXiv:2610.02632},
howpublished = {arXiv},
url = {https://arxiv.org/abs/2610.02632},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_f13c4495cd4b; sha256 8b3bf5e31b07afee2b3795c0fc9318c68c5c073eb7d67112ec94cf797b1d9414},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.02632. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
As large language models are increasingly deployed for multi-step reasoning, verifying the correctness of their outputs has become essential for maintaining reliability at scale. Verifying the correctness of large language model outputs is often done by querying a costly ground-truth oracle, which is impractical to invoke at every step in an online setting. Prior work addresses this by querying a single weak verifier on every step, and using its score to decide whether the costly strong verifier needs to be queried as well, reserving strong verification for only a small fraction of the steps. However, a single fixed weak verifier may not perform consistently well as the subject matter or difficulty of incoming queries changes over time, and committing to one in advance risks either overly costly or inaccurate verification. We introduce OMVV (Online Multi-Verifier Verification), an algorithm that maintains a pool of $K$ candidate weak verifiers with differing cost and verification performance, and adaptively routes each round's decision to a verifier selected via an online score combiner and an exponential-weights routing policy. OMVV provides a distribution-free, finite-time guarantee on false-accept and false-reject rates across the full pool of verifiers, and further achieves sublinear regret against the best fixed verifier in hindsight under a combined cost and consistency objective. Experiments on reasoning dataset benchmarks show that OMVV achieves higher accuracy at lower verification cost than any single fixed verifier, across a range of operating budgets. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2610.02632 — 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 2026-10-05T23:01:38.615Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).