Predicting magnetism with first-principles AI

TL;DR

This paper introduces a neural-network variational Monte Carlo method to predict magnetic properties of materials directly from the many-electron Schrödinger equation, enabling efficient and accurate discovery of magnetic materials.

Contribution

It demonstrates the application of neural-network variational Monte Carlo to predict magnetism in complex materials without prior physics input, reducing computational cost and improving reliability.

Findings

Predicted itinerant ferromagnetism in WSe₂/WS₂

Identified antiferromagnetic insulator in twisted Γ-valley homobilayer

Achieved magnetic state predictions from a single S_z=0 calculation

Abstract

Computational discovery of magnetic materials remains challenging because magnetism arises from the competition between kinetic energy and Coulomb interaction that is often beyond the reach of standard electronic-structure methods. Here we tackle this challenge by directly solving the many-electron Schr\"odinger equation with neural-network variational Monte Carlo, which provides a highly expressive variational wavefunction for strongly correlated systems. Applying this technique to transition metal dichalcogenide moir\'e semicondutors, we predict itinerant ferromagnetism in WSe$_2$/WS$_2$ and an antiferromagnetic insulator in twisted $\Gamma$-valley homobilayer, using the same neural network without any physics input beyond the microscopic Hamiltonian. Crucially, both types of magnetic states are obtained from a single calculation within the $S_z=0$ sector, removing the need to compute and compare multiple $S_z$ sectors. This significantly reduces computational cost and paves the way for faster and more reliable magnetic material design.