• Event Date: April 30, 2026
  • Event Start Time: 12:10 PM
  • Event End Time: 1:10 PM
  • Event Type: Mathematical Physics In Person Seminar
  • Event Location: Hill 705

Luc Rey-Bellet – UMASS

Date/Time/Location


Thursday,
April 30, 2026, 12:10 pm; Hill Center 705


Proximal optimal transport divergences for generative modeling and sampling.

The motivation for this work comes from generative modeling and sampling problems where the target distributions are either supported on low-dimensional structure and/or known only through empirical samples. 

Through proximal regularization we construct new information theoretic divergences which combine in a flexible manner the desirable properties of relative entropy  and optimal transport (Wasserstein distances).  Using these divergences, associated Wasserstein gradient flows, and a mixture of implicit and explicit schemes we build efficient sampling and generative modeling algorithms.