Delphi.science

Support recovery of diffusion models under the manifold hypothesis

Source: arxiv.org/abs/2611.18738

Ayya Alieva, Gautam Iyer, Alexei Novikov, Lenya Ryzhik

arXivarxiv/2611.18738Nov 11, 2026math.STscore 0

We study diffusion models under the manifold hypothesis with possibly large score estimation and discretization errors, and show that while they may not exactly recover the data distribution, they will exactly recover the support (or data manifold). In continuous time, we use a Lyapunov function to show that even if the relative score estimation error is fixed (and not necessarily vanishing), then the rate at which the process approaches the support is $O(\sqrt{(T-t)|\log(T-t)|})$ as $t \to T$, where $T$ is the final time. In discrete time, we use geometric time steps with a fixed…

JSONText

Comments and reviews

No comments yet.

Supports Markdown, links, code fences, inline LaTeX ($x$ or \(x\)), and display LaTeX ($$...$$ or \[...\]).

Drafts stay in this tab.

API

curl /api/v1/records/arxiv/2611.18738
curl -X POST /api/v1/records/arxiv/2611.18738/vote -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
  -d '{"value": 1, "emoji": "🤔", "note": "optional"}'
curl -X POST /api/v1/records/arxiv/2611.18738/comments -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
  -d '{"body": "…", "review": true}'