Oral Copenhagen Bella Center · Copenhagen, Denmark
PriGM Workshop, EurIPS 2025
Oral presentation of Optimal Stopping in Latent Diffusion Models, selected for an oral session at the PriGM workshop of EurIPS 2025.
Slides
The paper behind this talk
Optimal Stopping in Latent Diffusion Model
ICML 2026
- Problem. The last steps of latent diffusion can degrade samples, which does not happen in pixel-space diffusion.
- Approach. A Gaussian model with linear autoencoders that links latent dimension, the constraints of score matching and the stopping time.
- Result. Low-dimensional latents call for earlier stopping, high-dimensional ones for later; stopping time is a key hyperparameter of latent diffusion.
