The paper investigates score‑matching diffusion models that use a convolutional architecture, arguing that their locality bias allows effective field theory from physics to describe denoising dynamics. The authors first apply this framework to a toy example with an analytical solution, then to MNIST, demonstrating that in both settings the mutual information between two points increases as predicted by a simple effective field theory of Brownian motion.
arXiv:2409. 18804v3 Announce Type: replace-cross Abstract: Denoising Diffusion Probabilistic Models (DDPM) are powerful state-of-the-art methods used to generate synthetic data from high-dimensional data distributions and are widely used for image, audio, and video generation as well as many more applications in science and beyond.
arXiv:2603. 22962v3 Announce Type: replace Abstract: We study the theoretical behavior of denoising score matching--the learning task associated to diffusion models--when the data distribution is supported on a low-dimensional manifold and the score is parameterized using a random feature neural network.
arXiv:2606. 15835v1 Announce Type: cross Abstract: Diffusion models have achieved impressive empirical success in generative tasks, and their convergence theory is now relatively well understood.
arXiv:2606. 19894v1 Announce Type: new Abstract: The remarkable success of score-based diffusion models has spurred significant efforts to establish their theoretical foundations.
arXiv:2605. 22950v2 Announce Type: replace-cross Abstract: Score matching is an alternative to maximum likelihood estimation when the normalizing constant is unknown or too costly to evaluate.
arXiv:2609. 04822v1 Announce Type: cross Abstract: While diffusion-based methods have recently emerged as effective tools for probing the intrinsic geometry of high-dimensional data, their statistical difficulty remains largely unexplored.
arXiv:2509. 23937v5 Announce Type: replace-cross Abstract: Diffusion models transform noise into data by injecting information that was captured in their neural network during the training phase.
arXiv:2606. 06179v1 Announce Type: cross Abstract: Score-based diffusion models are typically trained by minimizing the $L^2$ score matching error, and standard theoretical analyses rely on this quantity to bound the sampling discrepancy between the learned and target distributions.
arXiv:2607. 01693v1 Announce Type: new Abstract: These notes give a proof-oriented introduction to diffusion models from the viewpoint of sampling, tracing a single arc from classical sampling dynamics to modern diffusion samplers, their error analysis, and inference-time control.