score-based generative models
A class of generative models that utilize a score function to guide the sampling of data points, optimizing the generation process based on gradient estimates of data distributions, employed in tasks like image synthesis and complex data generation.
- Adversary Aware Optimization for Robust Defense
- Algorithm- and Data-Dependent Generalization Bounds for Diffusion Models
- Approximation and Generalization Abilities of Score-based Neural Network Generative Models for Sub-Gaussian Distributions
- Preconditioned Langevin Dynamics with Score-based Generative Models for Infinite-Dimensional Linear Bayesian Inverse Problems
- System-Embedded Diffusion Bridge Models
- Unified all-atom molecule generation with neural fields
- Wasserstein Convergence of Critically Damped Langevin Diffusions