REVIEW 9 cited by
Towards Faster Non-Asymptotic Convergence for Diffusion-Based Generative Models
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Towards Faster Non-Asymptotic Convergence for Diffusion-Based Generative Models
read the original abstract
Diffusion models, which convert noise into new data instances by learning to reverse a Markov diffusion process, have become a cornerstone in contemporary generative modeling. While their practical power has now been widely recognized, the theoretical underpinnings remain far from mature. In this work, we develop a suite of non-asymptotic theory towards understanding the data generation process of diffusion models in discrete time, assuming access to $\ell_2$-accurate estimates of the (Stein) score functions. For a popular deterministic sampler (based on the probability flow ODE), we establish a convergence rate proportional to $1/T$ (with $T$ the total number of steps), improving upon past results; for another mainstream stochastic sampler (i.e., a type of the denoising diffusion probabilistic model), we derive a convergence rate proportional to $1/\sqrt{T}$, matching the state-of-the-art theory. Imposing only minimal assumptions on the target data distribution (e.g., no smoothness assumption is imposed), our results characterize how $\ell_2$ score estimation errors affect the quality of the data generation processes. In contrast to prior works, our theory is developed based on an elementary yet versatile non-asymptotic approach without resorting to toolboxes for SDEs and ODEs. Further, we design two accelerated variants, improving the convergence to $1/T^2$ for the ODE-based sampler and $1/T$ for the DDPM-type sampler, which might be of independent theoretical and empirical interest.
Forward citations
Cited by 9 Pith papers
-
Intrinsic Wasserstein Rates for Score-Based Generative Models on Smooth Manifolds
Score-based generative models attain intrinsic Wasserstein-1 sample rates of order n to the power of -(beta+1)/(d+2beta) on d-dimensional smooth manifolds with beta-Holder densities.
-
Proximal-Based Generative Modeling for Bayesian Inverse Problems
PGM framework links diffusion to proximal regularization for closed-form Moreau-score sampling in Bayesian inverse problems, learned only from prior samples.
-
Proximal-Based Generative Modeling for Bayesian Inverse Problems
PGM replaces the intractable likelihood score in diffusion models with a closed-form Moreau score computed via proximal operators, enabling non-asymptotic sampling for inverse problems trained only on prior data.
-
Tessellations of Semi-Discrete Flow Matching
Semi-discrete Flow Matching produces terminal assignment regions that are topologically simple (open, simply connected, homeomorphic to the ball under assumption) yet geometrically distinct from optimal transport Lagu...
-
From Score Matching to Diffusion: A Fine-Grained Error Analysis in the Gaussian Setting
In the Gaussian setting the Wasserstein error of score-matching-plus-diffusion sampling equals a kernel norm of the data power spectrum whose kernel is determined by the four error sources and the algorithm parameters.
-
Diffusion Models Adapt to Low-Dimensional Structure Under Flexible Coefficient Choices
For a broad class of coefficients, diffusion models achieve Õ(k/ε) iteration complexity for ε-accurate TV sampling under low-dimensional structure, independent of ambient dimension.
-
The Value of Covariance Matching in Gaussian DDPMs and the Lanczos Sampler
Full covariance matching in Gaussian DDPMs yields O(1/T^2) path KL error and is enabled by the training-free Lanczos Gaussian sampler using Jacobian-vector products.
-
Energy Generative Modeling: A Lyapunov-based Energy Matching Perspective
Training and sampling in static scalar energy generative models are two instances of the same Lyapunov-driven density transport dynamics on Wasserstein space, differing only by initial condition, which yields a finite...
-
HYVINT: Intensity-Driven Hypergraph Generation with Variational Representations
HYVINT introduces an intensity-driven incidence mechanism and tractable variational estimator for hypergraph generation, with error bounds and empirical gains in fidelity, novelty, and diversity.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.