English
Related papers

Related papers: Steering Dynamical Regimes of Diffusion Models by …

200 papers

Diffusion based generative models have achieved unprecedented fidelity in synthesizing high dimensional data, yet the theoretical mechanisms governing multimodal generation remain poorly understood. Here, we present a theoretical framework…

Machine Learning · Computer Science 2026-02-10 Emil Albrychiewicz , Andrés Franco Valiente , Li-Ching Chen

Using statistical physics methods, we study generative diffusion models in the regime where the dimension of space and the number of data are large, and the score function has been trained optimally. Our analysis reveals three distinct…

Machine Learning · Computer Science 2025-01-08 Giulio Biroli , Tony Bonnaire , Valentin de Bortoli , Marc Mézard

It is known that the distribution of nonreversible Markov processes breaking the detailed balance condition converges faster to the stationary distribution compared to reversible processes having the same stationary distribution. This is…

Statistical Mechanics · Physics 2021-06-30 Francesco Coghi , Raphael Chetrite , Hugo Touchette

Diffusion models generate high-dimensional data such as images by learning a process that gradually removes noise from corrupted data. Recent studies have shown that the backward dynamics of diffusion models exhibit two characteristic…

Statistical Mechanics · Physics 2026-04-14 Tomoei Takahashi , Takashi Takahashi , Yoshiyuki Kabashima

As a generalization of deterministic, nonlinear conservative dynamical systems, a notion of {\em canonical conservative dynamics} with respect to a positive, differentiable stationary density $\rho(x)$ is introduced: $\dot{x}=j(x)$ in which…

Mathematical Physics · Physics 2013-05-09 Hong Qian

In this paper, we analyze the use of the Ornstein-Uhlenbeck process to model dynamical systems subjected to bounded noisy perturbations. In order to discuss the main characteristics of this new approach we consider some basic models in…

Dynamical Systems · Mathematics 2024-01-17 Tomás Caraballo , Renato Colucci , Javier López-de-la-Cruz , Alain Rapaport

Generative diffusion models have recently emerged as a leading approach for generating high-dimensional data. In this paper, we show that the dynamics of these models exhibit a spontaneous symmetry breaking that divides the generative…

Machine Learning · Computer Science 2023-10-27 Gabriel Raya , Luca Ambrogioni

Inference-time steering enables pretrained diffusion/flow models to be adapted to new tasks without retraining. A widely used approach is the ratio-of-densities method, which defines a time-indexed target path by reweighting…

Artificial Intelligence · Computer Science 2025-12-12 Ziseok Lee , Minyeong Hwang , Sanghyun Jo , Wooyeol Lee , Jihyung Ko , Young Bin Park , Jae-Mun Choi , Eunho Yang , Kyungsu Kim

Diffusion models have become the de facto framework for generating new datasets. The core of these models lies in the ability to reverse a diffusion process in time. The goal of this manuscript is to explain, from a PDE perspective, how…

Probability · Mathematics 2025-01-29 Fei Cao , Kimball Johnston , Thomas Laurent , Justin Le , Sébastien Motsch

We analyze the time reversed dynamics of generative diffusion models. If the exact empirical score function is used in a regime of large dimension and exponentially large number of samples, these models are known to undergo transitions…

Statistics Theory · Mathematics 2025-11-17 Anand Jerry George , Rodrigo Veiga , Nicolas Macris

Guiding unconditional diffusion models typically requires either retraining with conditional inputs or per-step gradient computations (e.g., classifier-based guidance), both of which incur substantial computational overhead. We present a…

Machine Learning · Computer Science 2026-02-13 Qingsong Wang , Mikhail Belkin , Yusu Wang

We explore the dynamics of active elements performing persistent random motion with fluctuating active speed and in the presence of translational noise in a $d$-dimensional harmonic trap, modeling active speed generation through an…

Statistical Mechanics · Physics 2025-02-18 Manish Patel , Amir Shee , Debasish Chaudhuri

A central problem in machine learning involves modeling complex data-sets using highly flexible families of probability distributions in which learning, sampling, inference, and evaluation are still analytically or computationally…

Machine Learning · Computer Science 2015-11-20 Jascha Sohl-Dickstein , Eric A. Weiss , Niru Maheswaranathan , Surya Ganguli

We study the first-passage dynamics of a non-Markovian stochastic process with time-averaged feedback, which we model as a one-dimensional Ornstein--Uhlenbeck process wherein the particle drift is modified by the empirical mean of its…

Statistical Mechanics · Physics 2025-09-16 Francesco Coghi , Romain Duvezin , John S. Wettlaufer

Generative modeling via stochastic processes has led to remarkable empirical results as well as to recent advances in their theoretical understanding. In principle, both space and time of the processes can be discrete or continuous. In this…

Machine Learning · Statistics 2024-05-07 Ludwig Winkler , Lorenz Richter , Manfred Opper

Diffusion models generate high-quality synthetic data. They operate by defining a continuous-time forward process which gradually adds Gaussian noise to data until fully corrupted. The corresponding reverse process progressively "denoises"…

This work is concerned with model reduction of stochastic differential equations and builds on the idea of replacing drift and noise coefficients of preselected relevant, e.g. slow variables by their conditional expectations. We extend…

Analysis of PDEs · Mathematics 2020-03-05 Carsten Hartmann , Lara Neureither , Upanshu Sharma

We solve a physically significant extension of a classic problem in the theory of diffusion, namely the Ornstein-Uhlenbeck process [G. E. Ornstein and L. S. Uhlenbeck, Phys. Rev. 36, 823, (1930)]. Our generalised Ornstein-Uhlenbeck systems…

Statistical Mechanics · Physics 2009-11-11 V. Bezuglyy , B. Mehlig , M. Wilkinson , K. Nakamura , E. Arvedson

Unbalanced probability circulation, which yields cyclic motions in phase space, is the defining characteristics of a stationary diffusion process without detailed balance. In over-damped soft matter systems, such behavior is a hallmark of…

Mathematical Physics · Physics 2015-09-22 Hong Qian

Diffusion models generate samples by incrementally reversing a process that turns data into noise. We show that when the step size goes to zero, the reversed process is invariant to the distribution of these increments. This reveals a…

Machine Learning · Computer Science 2024-12-12 Henry Li
‹ Prev 1 2 3 10 Next ›