Metraipolis › archive › dc_86724afe7afe

Estimating Population-Risk Curves Along Nonconvex Gradient Flows from the Training Sample — arXiv:2608.30261

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.30261
License (as archived)
CC0-1.0
Archived
2026-10-17T08:39:57.367Z
SHA-256 of the archived text
fd50a2b28e8e377582c7ad1520f69e72308a9bd1489b51d4990adbd364c7b323
Stable URL of this page
https://metraipolis.com/documento/dc_86724afe7afe

How to cite

Estimating Population-Risk Curves Along Nonconvex Gradient Flows from the Training Sample — arXiv:2608.30261 — arXiv (en), https://arxiv.org/abs/2608.30261 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_86724afe7afe (Metraipolis archive dc_86724afe7afe, sha256 fd50a2b28e8e377582c7ad1520f69e72308a9bd1489b51d4990adbd364c7b323)
BibTeX
@misc{dc86724afe7afe,
  title = {Estimating Population-Risk Curves Along Nonconvex Gradient Flows from the Training Sample — arXiv:2608.30261},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.30261},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_86724afe7afe; sha256 fd50a2b28e8e377582c7ad1520f69e72308a9bd1489b51d4990adbd364c7b323},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.30261. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We estimate the conditional population-risk curve of a realized smooth nonconvex gradient flow from the training sample. Flow approximate leave-one-out (Flow-ALO) propagates a deletion response and evaluates omitted observations at approximate deleted paths. The risk-curve error decomposes into response approximation, exact-LOO fluctuation, and deletion-to-full risk transfer. On each fixed finite horizon, bounded centered training-loss gradients, a one-sided Hessian lower bound, locally Lipschitz Hessians, and a strict tube-closure condition yield an explicit $(n-1)^{-2}$ bound for the deletion-response error. Bounded evaluation-loss gradients transfer the deletion-response bound to the score without requiring the Hessian to be invertible. Direct first-order jackknife cancellation and exact-LOO concentration control deletion-to-full risk transfer and fluctuation, respectively, completing recovery of the conditional population-risk curve. For bounded smooth two-layer mean-field networks training both layers, the score-error bound is uniform in width.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-01 in cs.LG).
Categorie arXiv: Machine Learning (stat.ML) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2608.30261 — 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 2026-10-17T08:13:26.934Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).