Opportunities and challenges of diffusion models for generative AI

M Chen, S Mei, J Fan, M Wang - National Science Review, 2024 - academic.oup.com
Diffusion models, a powerful and universal generative artificial intelligence technology, have
achieved tremendous success and opened up new possibilities in diverse applications. In …

Dynamical regimes of diffusion models

G Biroli, T Bonnaire, V De Bortoli, M Mézard - Nature Communications, 2024 - nature.com
We study generative diffusion models in the regime where both the data dimension and the
sample size are large, and the score function is trained optimally. Using statistical physics …

An overview of diffusion models: Applications, guided generation, statistical rates and optimization

M Chen, S Mei, J Fan, M Wang - ar**: finite fisher information is all you need
G Conforti, A Durmus, M Gentiloni Silveri - arxiv e-prints, 2023 - ui.adsabs.harvard.edu
Diffusion models are a new class of generative models that revolve around the estimation of
the score function associated with a stochastic differential equation. Subsequent to its …

Stochastic runge-kutta methods: Provable acceleration of diffusion models

Y Wu, Y Chen, Y Wei - arxiv preprint arxiv:2410.04760, 2024 - arxiv.org
Diffusion models play a pivotal role in contemporary generative modeling, claiming state-of-
the-art performance across various domains. Despite their superior sample quality …

Score-based Diffusion Models via Stochastic Differential Equations--a Technical Tutorial

W Tang, H Zhao - arxiv preprint arxiv:2402.07487, 2024 - arxiv.org
This is an expository article on the score-based diffusion models, with a particular focus on
the formulation via stochastic differential equations (SDE). After a gentle introduction, we …

Denoising diffusion probabilistic models are optimally adaptive to unknown low dimensionality

Z Huang, Y Wei, Y Chen - arxiv preprint arxiv:2410.18784, 2024 - arxiv.org
The denoising diffusion probabilistic model (DDPM) has emerged as a mainstream
generative model in generative AI. While sharp convergence guarantees have been …