stat.MLJun 26, 2025

Critically-Damped Higher-Order Langevin Dynamics for Generative Modeling

Authors: Benjamin Sterling, Chad Gueli, Mónica F. Bugallo

Abstract

Denoising diffusion probabilistic models (DDPMs) represent an entirely new class of generative AI methods that have yet to be fully explored. They use Langevin dynamics, represented as stochastic differential equations, to describe a process that transforms data into noise, the forward process, and a process that transforms noise into generated data, the reverse process. Many of these methods utilize auxiliary variables that formulate the data as a position" variable, and the auxiliary variables are referred to as velocity", acceleration", etc. In this sense, it is possible to critically damp" the dynamics. Critical damping has been successfully introduced in Critically-Damped Langevin Dynamics (CLD) and Critically-Damped Third-Order Langevin Dynamics (TOLD++), but has not yet been applied to dynamics of arbitrary order. The proposed methodology generalizes Higher-Order Langevin Dynamics (HOLD), a recent state-of-the-art diffusion method, by introducing the concept of critical damping from systems analysis. Similarly to TOLD++, this work proposes an optimal set of hyperparameters in the nn-dimensional case, where HOLD leaves these to be user defined. Additionally, this work provides closed-form solutions for the mean and covariance of the forward process that greatly simplify its implementation. Experiments are performed on the CIFAR-10 and CelebA-HQ 256×256256 \times 256 datasets, and validated against the FID metric.

Figures & tables

Appendix figures & tables1 asset

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. Reducing Diffusion Model Memorization with Higher Order Langevin Dynamics

    May 18, 2026Benjamin Sterling, Mónica F. Bugallo, Tom TirerLangevin DynamicsMemorization in Generative Models

  2. Annealed Lévy Diffusion via Gaussian Surrogate Reverse Dynamics

    Oct 14, 2026Yuling Jiao, Lican Kang, Jerry Zhijian Yang +2Diffusion SamplingDiffusion Models

  3. Slowly Annealed Langevin Dynamics: Theory and Applications to Training-Free Guided Generation

    May 8, 2026Atsushi Nitanda, Dake Bu, Yueming Lyu +1Diffusion Model GuidanceLangevin Dynamics