[论文] Quenched Ensemble Sampling

论文概要

研究领域: ML
作者: David Yallup
发布时间: 2026-09-14
arXiv: 2609.15894

中文摘要

物理系统能量函数采样中最尖锐的挑战出现在相变时——状态密度突变，许多采样算法停滞。嵌套采样是一种在硬能量约束下穿越状态密度的粒子方法，已知对此类相变具有鲁棒性，但其在高维中的应用受限于在该约束下采样的困难。本文引入淬灭集成采样（Quenched Ensemble Sampling），将硬约束推广为能量边界处的一族排斥势。这保留了能量单调递减的淬灭路径，同时使约束目标适用于可扩展的基于梯度的核。我们在相变合成模型上证明，该方法在一级相变处估计边缘似然并抽取后验样本，而回火等流行替代方法则失败。我们将该过程应用于贝叶斯神经网络的边缘似然估计，实现了网络架构之间的模型比较。最后，在一个高维连续晶格场论中，我们展示了该方法穿越一级相变并估计配分函数。

原文摘要

Some of the sharpest challenges in sampling from the energy functions of physical systems arise at phase transitions, where the density of states changes abruptly and many sampling algorithms stall. Nested sampling is a particle method that traverses the density of states under a hard energy constraint and is known to be robust to such transitions, but its application in high dimension is limited by the difficulty of sampling under that constraint. In this work we introduce Quenched Ensemble Sampling, which generalises the hard constraint to a family of repulsive potentials at the energy boundary. This preserves the quenched path of monotonically decreasing energy while making the constrained target amenable to scalable gradient-based kernels. We demonstrate on synthetic models of phase tr...

自动采集于 2026-09-16

#论文 #arXiv #ML #小凯