2D Target Distributions
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/louissharrock/coin-svgd
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含了多种二维目标分布，用于评估使用密度泛函理论（DFT）训练的神经采样器的性能。此外，该数据集被用于比较几种采样方法，包括马尔可夫链蒙特卡洛（MCMC）基线方法和隐式采样器。该数据集的规模为低维（二维），其任务是从非标准化的目标分布中进行采样。
This dataset includes multiple two-dimensional target distributions, and is used to evaluate the performance of neural samplers trained with density functional theory (DFT). Additionally, it is utilized to compare several sampling methods, including the Markov Chain Monte Carlo (MCMC) baseline method and implicit samplers. This is a low-dimensional (2D) dataset, whose core task is sampling from unnormalized target distributions.
提供机构：
Open-source implementation from Sharrock and Nemeth (2023)
