paper-with-me

홈 › Papers

One-Step Sampler for Boltzmann Distributions via Drifting

2026-03-18 · Wenhan Cao, Keyu Yan, Lin Zhao arxiv

We present a drifting-based framework for amortized sampling of Boltzmann distributions defined by energy functions. The method trains a one-step neural generator by projecting samples along a Gaussian-smoothed score field from the current model distribution toward the target Boltzmann distribution. For targets specified only up to an unknown normalization constant, we derive a practical target-side drift from a smoothed energy and use two estimators: a local importance-sampling mean-shift estimator and a second-order curvature-corrected approximation. Combined with a mini-batch Gaussian mean-shift estimate of the sampler-side smoothed score, this yields a simple stop-gradient objective for stable one-step training. On a four-mode Gaussian-mixture Boltzmann target, our sampler achieves mean error $0.0754$, covariance error $0.0425$, and RBF MMD $0.0020$. Additional double-well and banana targets show that the same formulation also handles nonconvex and curved low-energy geometries. Overall, the results support drifting as an effective way to amortize iterative sampling from Boltzmann distributions into a single forward pass at test time.

📄 PDF Abstract BibTeX arXiv:2603.17579

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Training Neural Samplers with Reverse Diffusive KL Divergence

2024-10-16 · Jiajun He, Wenlin Chen, Mingtian Zhang, David Barber 외

Training generative models to sample from unnormalized density functions is an important and challenging task in machine learning. Traditional training methods often rely on the reverse Kullback-Leibler (KL) divergence d…

A Langevin-like Sampler for Discrete Distributions

2022-06-20 · Ruqi Zhang, Xingchao Liu, Qiang Liu

We propose discrete Langevin proposal (DLP), a simple and scalable gradient-based proposal for sampling complex high-dimensional discrete distributions. In contrast to Gibbs sampling-based methods, DLP is able to update …

Efficient ExplorationText Generation

Pseudo-extended Markov chain Monte Carlo

2017-08-17 · NeurIPS 2019 12 · Christopher Nemeth, Fredrik Lindsten, Maurizio Filippone, James Hensman

Sampling from posterior distributions using Markov chain Monte Carlo (MCMC) methods can require an exhaustive number of iterations, particularly when the posterior is multi-modal as the MCMC sampler can become trapped in…

Lookahead Drifting Model

2026-04-10 · Guoqiang Zhang, Kenta Niwa, W. Bastiaan Kleijn arxiv

Recently, a new paradigm named \emph{drifting model} has been proposed for mapping distributions, which achieves the SOTA image generation performance over ImageNet via one-step neural functional evaluation (NFE). The ba…

Image Generation

Oops I Took A Gradient: Scalable Sampling for Discrete Distributions

2021-02-08 · Will Grathwohl, Kevin Swersky, Milad Hashemi, David Duvenaud 외

We propose a general and scalable approximate sampling strategy for probabilistic models with discrete variables. Our approach uses gradients of the likelihood function with respect to its discrete inputs to propose upda…