Metraipolis › archive › dc_0ec241d3fd58
Published Unlearning Numbers Move Per Checkpoint, and Not Because the Removed Data Survives: An Audit of 263 Released Batch-Normalized Checkpoints — arXiv:2609.11490
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.11490
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-16T05:24:23.918Z
- SHA-256 of the archived text
9c2c98728ffdabf31408592150e94b59fe72603af73f98f1dd3227ad2d419e4d- Stable URL of this page
https://metraipolis.com/documento/dc_0ec241d3fd58
How to cite
Published Unlearning Numbers Move Per Checkpoint, and Not Because the Removed Data Survives: An Audit of 263 Released Batch-Normalized Checkpoints — arXiv:2609.11490 — arXiv (en), https://arxiv.org/abs/2609.11490 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0ec241d3fd58 (Metraipolis archive dc_0ec241d3fd58, sha256 9c2c98728ffdabf31408592150e94b59fe72603af73f98f1dd3227ad2d419e4d)
BibTeX
@misc{dc0ec241d3fd58,
title = {Published Unlearning Numbers Move Per Checkpoint, and Not Because the Removed Data Survives: An Audit of 263 Released Batch-Normalized Checkpoints — arXiv:2609.11490},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.11490},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0ec241d3fd58; sha256 9c2c98728ffdabf31408592150e94b59fe72603af73f98f1dd3227ad2d419e4d},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.11490. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
An unlearning audit reads its verdict off numbers that an unlearned model and its retrained reference each publish, and both also ship batch-normalization statistics that no gradient step wrote and no release records. Refitting them on kept data at bit-identical weights moves 47 of 221 released checkpoints past the spread their own release's seeds show, several inside a method whose average does not move: what moves is the checkpoint's property, not its method's. What does the moving is not the removed data surviving in the state: exchanging kept records for removed ones inside a fixed fitting pool moves a published cell by almost nothing, while how far a checkpoint's shipped state has drifted from any refit does track it. The consequence for a published decision is real but narrow: twelve verdicts cross, four clear a measured recalibration budget, two clear it on every replicate, and a population we trained and sited near its own criterion yields none. A release should therefore name the fitting convention beside the number, on the batch-normalized vision models where this channel exists. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-11 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI) ; Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2609.11490 — 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 2026-10-16T05:17:52.027Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-11?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).