Metraipolis › archive › dc_385bd4443084
Certifying Model Upgrades with Slice-Wise Non-Regression and Incumbent Fallback — arXiv:2609.13714
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.13714
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-16T00:10:27.470Z
- SHA-256 of the archived text
dc5511d6015ceec09008e38a2c4d33330c42995839432de15143e9c5e8480063- Stable URL of this page
https://metraipolis.com/documento/dc_385bd4443084
How to cite
Certifying Model Upgrades with Slice-Wise Non-Regression and Incumbent Fallback — arXiv:2609.13714 — arXiv (en), https://arxiv.org/abs/2609.13714 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_385bd4443084 (Metraipolis archive dc_385bd4443084, sha256 dc5511d6015ceec09008e38a2c4d33330c42995839432de15143e9c5e8480063)
BibTeX
@misc{dc385bd4443084,
title = {Certifying Model Upgrades with Slice-Wise Non-Regression and Incumbent Fallback — arXiv:2609.13714},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.13714},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_385bd4443084; sha256 dc5511d6015ceec09008e38a2c4d33330c42995839432de15143e9c5e8480063},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.13714. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
An updated model can improve an aggregate metric while degrading a slice that matters to a downstream user. We study checkpoint selection subject to non-regression tolerances relative to a retained incumbent. The central distinction is between failing to detect harm and certifying non-inferiority: the former can release harmful updates with high probability when evaluation is noisy. We give a reproducible release procedure that separates candidate search from independent, paired evaluation and returns the exact incumbent when certification fails. Applying established intersection-union and Learn-then-Test principles, we state finite-sample guarantees for one frozen candidate, a finite candidate library, and a prespecified testing order. A joint release decision does not require a slice-count Bonferroni penalty, although certification power can still decrease with the number of slices. In bounded-score simulations, a no-detected-harm gate releases a harmful candidate in 99.7% of trials in one 32-slice setting, compared with 2.6% for an exact non-inferiority gate at a 5% target. A constructed two-block family yields larger certified utility than a scalar path under matched candidate counts. Public digits experiments, including a subsequent continuation that improves average aggregate accuracy, return the incumbent in every run because certification is underpowered. These results establish an auditable protocol and its limitations; they do not establish benefits on foundation-model or multilingual translation upgrades. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-15 in cs.LG). Categorie arXiv: Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2609.13714 — 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 2026-10-15T23:54:14.632Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-15?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).