Abstract:Metallic spin glass systems, such as dilute magnetic alloys, are characterized by randomly distributed local moments coupled to each other through a long-range electron-mediated effective interaction. We present a scalable machine learning (ML) framework for dynamical simulations of metallic spin glasses. A Behler-Parrinello type neural-network model, based on the principle of locality, is developed to accurately and efficiently predict electron-induced local magnetic fields that drive the spin dynamics. A crucial component of the ML model is a proper symmetry-invariant representation of local magnetic environment which is direct input to the neural net. We develop such a magnetic descriptor by incorporating the spin degrees of freedom into the atom-centered symmetry function methods which are widely used in ML force-field models for quantum molecular dynamics. We apply our approach to study the relaxation dynamics of an amorphous generalization of the s-d model. Our work highlights the promising potential of ML models for large-scale dynamical modeling of itinerant magnets with quenched disorder.

What problem does this paper attempt to address?

This paper aims to address the computational challenges in performing large-scale dynamical simulations of metallic spin glass systems. Spin glass is a disordered magnetic material with a random distribution of local magnetic moments that are coupled through long-range electron-mediated effective interactions. Traditional simulation methods require frequent solutions to the electronic Hamiltonian, resulting in high computational costs for large systems. The paper proposes a machine learning (ML)-based framework that adopts a Behler-Parrinello neural network model to predict the electron-induced local magnetic field, enabling efficient simulation of spin dynamics. The key is to develop a magnetic descriptor that preserves symmetry, taking into account the local environment of spin configurations and training it using neural networks. This approach allows for linearly scalable computational efficiency, similar to approximate methods in quantum molecular dynamics. Specifically, the paper first introduces a general model of metallic spin glasses based on the s-d model, where electron spins interact with local magnetic moments. It then discusses how, under the adiabatic approximation, electrons quickly reach a quasi-equilibrium state with spin configurations, allowing for the calculation of effective energy using the electron density operator. The paper then describes in detail how to construct the ML model, including using a magnetic version of atomic-centered symmetry function (ACSF) to generate symmetry-invariant feature variables, which are input into the neural network to predict local energy and magnetic field. Finally, the authors apply this ML force-field model to study the relaxation dynamics of a random s-d model, validating the accuracy and efficiency of the model by comparing it with exact solution benchmark datasets. This approach provides a new avenue for understanding and simulating large electronic spin systems with quenched disorder.

Machine learning force-field models for metallic spin glass

Learning Spin Liquids on a Honeycomb Lattice with Artificial Neural Networks.

Machine Learning Exchange Fields for Ab-initio Spin Dynamics

Machine learning atomic dynamics to unfold the origin of plasticity in metallic glasses: From thermo- to acousto-plastic flow

Machine learning force-field model for kinetic Monte Carlo simulations of itinerant Ising magnets

Coarsening of chiral domains in itinerant electron magnets: A machine learning force field approach

Numerical Simulations Of Random Spin (And Fermionic) Models With A Wide Distribution Of Energy Scales

Machine learning assisted derivation of minimal low-energy models for metallic magnets

Machine learning bridges local static structure with multiple properties in metallic glasses

Machine learning prediction of magnetic properties of Fe-based metallic glasses considering local structures

Machine learning atomic-scale stiffness in metallic glass

Machine learning versus human learning in predicting glass-forming ability of metallic glasses

Machine learning for phase ordering dynamics of charge density waves

On Machine Learning Force Fields for Metallic Nanoparticles

Predicting the propensity for thermally activated β events in metallic glasses via interpretable machine learning

A transferable machine-learning framework linking interstice distribution and plastic heterogeneity in metallic glasses

Machine learning magnetism classifiers from atomic coordinates

Machine-Learning Force Fields for Metallic Materials: Phase Transformations and Deformations

Complex Spin Hamiltonian Represented by an Artificial Neural Network

Atomistic simulation assisted error-inclusive Bayesian machine learning for probabilistically unraveling the mechanical properties of solidified metals

Kinetics of orbital ordering in cooperative Jahn-Teller models: Machine-learning enabled large-scale simulations