Reproduce “Sampling the XY model with an attention-based neural network” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Piotr Białas, Christian D. Morales, Tomasz Stebel, Dawid Zapolski

The authors extend tVAN to sample continuous spins in the two-dimensional XY model. It generates angles sequentially using von Mises mixtures, geometry-limited attention, and an approximate conditional based on neighboring spins. Importance reweighting estimates energy, absolute magnetization, and helicity modulus. At L=96, measurements agree with Wolff MCMC within statistical fluctuations at three temperatures; at beta 1.12, normalized ESS is 0.074316.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.