Metraipolis › archive › dc_d7c11fc020ef
Memorization and Malign Generalization in Conditional Diffusion Models with Random Features — arXiv:2610.11288
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.11288
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-09T06:23:15.756Z
- SHA-256 of the archived text
210164ef9be258e00c95de3bf3215a8cbe18b10bb1b1c66454273ab43c84ccf9- Stable URL of this page
https://metraipolis.com/documento/dc_d7c11fc020ef
How to cite
Memorization and Malign Generalization in Conditional Diffusion Models with Random Features — arXiv:2610.11288 — arXiv (en), https://arxiv.org/abs/2610.11288 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d7c11fc020ef (Metraipolis archive dc_d7c11fc020ef, sha256 210164ef9be258e00c95de3bf3215a8cbe18b10bb1b1c66454273ab43c84ccf9)
BibTeX
@misc{dcd7c11fc020ef,
title = {Memorization and Malign Generalization in Conditional Diffusion Models with Random Features — arXiv:2610.11288},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.11288},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d7c11fc020ef; sha256 210164ef9be258e00c95de3bf3215a8cbe18b10bb1b1c66454273ab43c84ccf9},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.11288. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Conditional diffusion models generate diverse, novel, and high-quality samples under prescribed conditions. However, theoretical understanding of their memorization and generalization remains limited, while recent works have characterized these behaviors primarily in unconditional settings. In this work, we analyze a random-feature conditional score model in the high-dimensional proportional limit, deriving asymptotic expressions for training and test losses. By decomposing the test loss, we show that in the overparameterized regime, increasing model width improves prediction of the condition-dependent mean while reducing within-condition prediction variance, a phenomenon we term "malign generalization." Furthermore, analyzing the training loss reveals that more informative conditions lead to memorization of training samples at smaller widths. These theoretical findings are supported by experiments with U-Net architectures on realistic data. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2610.11288 — 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 2026-10-09T05:50:31.204Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).