Diffusion model tutorial

Add to your list(s) Download to your calendar using vCal

• Nikolay Malkin (University of Edinburgh)
• Monday 23 June 2025, 14:00-15:00
• Seminar Room 1, Newton Institute.

If you have a question about this talk, please contact nobody.

RCLW03 - Accelerating statistical inference and experimental design with machine learning

This tutorial is an introduction to diffusion models, a class of parametric generative processes that produces samples by iterative local refinement or denoising. In this tutorial, we will consider high-dimensional Euclidean spaces, although variants for discrete spaces and other Riemannian manifolds exist. The exposition will be self-contained assuming basic familiarity with latent variable models and principles of deep learning. I will present the main ideas from the point of view of hierarchical variational inference, with connections mentioned to stochastic differential equations and annealed Langevin dynamics. The talk will serve as background for my subsequent talk on diffusion modelling for amortised inference, i.e., in the absence of a ground-truth dataset.

This talk is part of the Isaac Newton Institute Seminar Series series.

This talk is included in these lists:

• All CMS events
• Featured lists
• INI info aggregator
• Isaac Newton Institute Seminar Series
• School of Physical Sciences
• Seminar Room 1, Newton Institute
• bld31

Note that ex-directory lists are not shown.