[PDF] arxiv.org

Zeroth-order sampling methods for non-log-concave distributions: Alleviating metastability by denoising diffusion

Y He, K Rojas, M Tao - arXiv preprint arXiv:2402.17886, 2024 - arxiv.org
This paper considers the problem of sampling from non-logconcave distribution, based on
queries of its unnormalized density. It first describes a framework, Diffusion Monte Carlo …
被引用次数：5 相关文章 所有 2 个版本
[PDF] arxiv.org

Provable Benefit of Annealed Langevin Monte Carlo for Non-log-concave Sampling

W Guo, M Tao, Y Chen - arXiv preprint arXiv:2407.16936, 2024 - arxiv.org
We address the outstanding problem of sampling from an unnormalized density that may be
non-log-concave and multimodal. To enhance the performance of simple Markov chain …
被引用次数：1 相关文章 所有 2 个版本
[PDF] arxiv.org

Faster Sampling via Stochastic Gradient Proximal Sampler

X Huang, D Zou, YA Ma, H Dong, T Zhang - arXiv preprint arXiv …, 2024 - arxiv.org
Stochastic gradients have been widely integrated into Langevin-based methods to improve
their scalability and efficiency in solving large-scale sampling problems. However, the …
被引用次数：1 相关文章 所有 3 个版本
[PDF] arxiv.org

An Improved Analysis of Langevin Algorithms with Prior Diffusion for Non-Log-Concave Sampling

X Huang, H Dong, D Zou, T Zhang - arXiv preprint arXiv:2403.06183, 2024 - arxiv.org
Understanding the dimension dependency of computational complexity in high-dimensional
sampling problem is a fundamental problem, both from a practical and theoretical …
相关文章 所有 2 个版本
[PDF] arxiv.org

A Practical Diffusion Path for Sampling

O Chehab, A Korba - arXiv preprint arXiv:2406.14040, 2024 - arxiv.org
Diffusion models are state-of-the-art methods in generative modeling when samples from a
target probability distribution are available, and can be efficiently sampled, using score …
相关文章