We’ve made progress towards stable and scalable training of energy-based models (EBMs) resulting in better sample quality and generalization ability than existing models. Generation in EBMs spends more compute to continually refine its answers and doing so can generate samples competitive with GANs at low temperatures, while also having mode coverage guarantees of likelihood-based models.
Energy-Based Models (EBMs) provide an interpretable framework for generative modeling of scientific data, but poor Markov Chain Monte Carlo mixing often limits their reliability. We introduce a training algorithm based on Parallel Trajectory Tempering (PTT), which exploits the continuity of the optimization path to maintain equilibrium sampling throughout learning.
arXiv:2607. 27077v1 Announce Type: new Abstract: Energy-Based Models (EBMs) provide an interpretable framework for generative modeling of scientific data, but poor Markov Chain Monte Carlo mixing often limits their reliability.
arXiv:2608. 07648v1 Announce Type: cross Abstract: Sampling high-dimensional probability distributions is a central task in scientific computing, with applications ranging from Bayesian inference to statistical physics and molecular simulation.
arXiv:2606. 23920v1 Announce Type: cross Abstract: The task of compositional generation involves using a conditional generative model, trained only on a subset of the possible conditions, to produce samples from compositionally-defined target distributions such as a geometric combination of the source distributions.
arXiv:2607. 01171v1 Announce Type: new Abstract: Sample-based generative models are increasingly used for probabilistic forecasting in high-stakes decision settings, yet their training objectives are blind to the decision maker's cost structure.
arXiv:2602. 03729v3 Announce Type: replace Abstract: Sampling from unnormalized probability densities is a central challenge in computational science.
arXiv:2605.29713v2 Announce Type: replace-cross Abstract: This book provides a compact, derivation-oriented introduction to the mathematical foundations of modern generative artificial intelligence....
The paper investigates memorization in diffusion models, showing that the empirical score function is a weighted sum of Gaussian score functions with sharp softmax weights, causing individual training samples to dominate and lead to sampling collapse. By approximating this function with a neural network, the authors obtain a smoother representation that generalizes better. They introduce two techniques—Noise Unconditioning and Temperature Smoothing—to further reduce single‑sample dominance, and demonstrate improved generalization across multiple datasets while preserving generation quality.
arXiv:2607. 19332v1 Announce Type: new Abstract: Generative models have undergone many generations of evolution, from VAEs/GANs to diffusion/flow matching.
The paper introduces a neural‑network method that learns a minimal‑deviation transformation of Monte Carlo simulated events to match one‑dimensional target distributions while preserving the multidimensional correlation structure of the original simulation. By operating under limited experimental information, the approach avoids the pitfalls of traditional one‑dimensional reweighting and the data‑hungry fully multidimensional corrections. Controlled pseudo‑data studies demonstrate improved agreement with target distributions and consistent multidimensional structure, making the method suitable for complex, high‑dimensional analyses where conventional techniques fall short.
The paper introduces a new method for perturbing data distributions in a way that respects equality constraints, allowing generative models to better handle constrained data. By adjusting the distribution while preserving the manifold geometry, the approach ensures support matches the ambient space dimension. Experiments with diffusion models and normalizing flows demonstrate improved data recovery and stable sampling across several tasks.