Metraipolis › archive › dc_d6433ef433db

Verification-Time Dependency on a Disappearing Evaluator — arXiv:2608.29912

Source: arXiv (en)
Original: https://arxiv.org/abs/2608.29912
License (as archived): CC0-1.0
Archived: 2026-10-17T08:19:19.423Z
SHA-256 of the archived text: fd9af6d5622888cc54f6d11504889eb62af49782aff8bc7ff0b9ac51dc202997
Stable URL of this page: https://metraipolis.com/documento/dc_d6433ef433db

How to cite

Verification-Time Dependency on a Disappearing Evaluator — arXiv:2608.29912 — arXiv (en), https://arxiv.org/abs/2608.29912 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d6433ef433db (Metraipolis archive dc_d6433ef433db, sha256 fd9af6d5622888cc54f6d11504889eb62af49782aff8bc7ff0b9ac51dc202997)

BibTeX

@misc{dcd6433ef433db,
  title = {Verification-Time Dependency on a Disappearing Evaluator — arXiv:2608.29912},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.29912},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d6433ef433db; sha256 fd9af6d5622888cc54f6d11504889eb62af49782aff8bc7ff0b9ac51dc202997},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.29912. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

AI governance and assurance often assume that a consequential model-mediated decision can be reconstructed or tested after the fact. That assumption may fail when the evaluator that produced the decision is no longer accessible in the same version and execution context. This paper develops three verification-time constructs derived from Execution Governance (EG) 3.0: Decision-State Commitment, Independent Verifiability, and Counterfactual Auditability. Independent reprocessing of released Study 2 artifacts reproduces two original within-family behavioural comparisons: 52.0% modal-decision reversal for Llama 3.1 8B versus Llama 3.3 70B (26/50) and 30.0% for GPT-OSS 20B versus GPT-OSS 120B (15/50). The corrected baseline establishes that these are within-family comparisons, not provider-established succession. Post-hoc re-pairing against Groq-designated migration paths yields 64.0% and 38.0% reversal, but these figures remain descriptive because the cross-family invocation parameters were asymmetric. A 22-event retirement census independently recomputes to median 16.45 months, mean 18.72 months, range 3.9-40.3 months, with 17/22 intervals below 24 months, while also showing that evaluator availability can differ by service surface. The joint contribution is an operational verification-time protocol and optional Verification-Time Preservation Package (VTPP) specifying what evidence to bind at authorization time, what a separately trusted verifier can substantiate later, how stability and paired counterfactual tests should be calibrated, and which semantic checks remain beyond JSON Schema validity. The protocol is downstream and non-authorizing: it does not alter the EG Core Formula, add a seventh live condition, or state jurisdiction-specific legal admissibility.

Tema degli esploratori: governance digitale e beni comuni (retro-catalogo arXiv: annunci del 2026-09-01 in cs.CY).
Categorie arXiv: Computers and Society (cs.CY) ; Software Engineering (cs.SE).
Pagina del lavoro: https://arxiv.org/abs/2608.29912 — 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 2026-10-17T08:07:56.355Z dalla pagina https://arxiv.org/catchup/cs.CY/2026-09-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).