Equivariant flow-based sampling for lattice gauge theory
Equivariant flow-based sampling for lattice gauge theory
复制标题
DOI：
10.1103/physrevlett.125.121601
复制
发表时间：
2020-03
影响因子：
8.6
通讯作者：
G. Kanwar;M. S. Albergo;D. Boyda;Kyle Cranmer;D. Hackett;S. Racanière;Danilo Jimenez Rezende;P. Shanahan
中科院分区：
文献类型：
--
作者：
G. Kanwar;M. S. Albergo;D. Boyda;Kyle Cranmer;D. Hackett;S. Racanière;Danilo Jimenez Rezende;P. Shanahan
我们定义了一类机器学习的基于流的采样算法的格规范理论，规范不变的建设。我们证明了这个框架的应用U（1）规范理论在两个时空维度，并发现，在小裸耦合，该方法是数量级更有效的采样拓扑量比更传统的采样程序，如混合蒙特卡罗和热浴。
We define a class of machine-learned flow-based sampling algorithms for lattice gauge theories that are gauge invariant by construction. We demonstrate the application of this framework to U(1) gauge theory in two spacetime dimensions, and find that, at small bare coupling, the approach is orders of magnitude more efficient at sampling topological quantities than more traditional sampling procedures such as hybrid Monte Carlo and heat bath.