Metraipolis › archive › dc_bb7e737ba0da

Post-Audit Mirage: Identifiability Limits of Offline Agent Verification — arXiv:2610.31130

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.31130
License (as archived)
CC0-1.0
Archived
2026-10-23T15:13:17.837Z
SHA-256 of the archived text
b75f2895d094be808dd7a2cae47070c2ab7c25779ebcc4de614c4dcb3539fc1d
Stable URL of this page
https://metraipolis.com/documento/dc_bb7e737ba0da

How to cite

Post-Audit Mirage: Identifiability Limits of Offline Agent Verification — arXiv:2610.31130 — arXiv (en), https://arxiv.org/abs/2610.31130 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_bb7e737ba0da (Metraipolis archive dc_bb7e737ba0da, sha256 b75f2895d094be808dd7a2cae47070c2ab7c25779ebcc4de614c4dcb3539fc1d)
BibTeX
@misc{dcbb7e737ba0da,
  title = {Post-Audit Mirage: Identifiability Limits of Offline Agent Verification — arXiv:2610.31130},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.31130},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_bb7e737ba0da; sha256 b75f2895d094be808dd7a2cae47070c2ab7c25779ebcc4de614c4dcb3539fc1d},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.31130. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Offline tests can fail to reveal whether an agent update will remain safe after deployment. We operationalize this post-audit mirage using matched safe and harmful lifecycles that expose identical verifier-visible evidence while an isolated scorer records opposite deployment truth. Our theorem applies standard observational-equivalence and indistinguishability ideas to this executable audit boundary and states the resulting error-abstention tradeoff. The Post-Audit Mirage benchmark makes that boundary executable across authorization, scheduling, and shared-capacity environments. IdentifiedRangeMonitor uses fresh group-labelled harm outcomes to return deploy, hold, cannot determine, or unsupported. Experiments show that monitoring can recover the missing distinction in favorable settings but may remain undecided for rare groups, small effects, or missing observations. Verifier-agent evaluation must therefore test both calculation validity and whether the available evidence supports the deployment decision.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.31130 — 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 2026-10-23T15:09:08.442Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).