Physics-Informed Weakly Supervised Learning for Interatomic Potentials

TL;DR

This paper introduces a physics-informed, weakly supervised learning method for interatomic potentials that enhances accuracy and robustness in atomistic simulations, especially with limited training data.

Contribution

It proposes two novel loss functions incorporating physical principles, improving MLIP training with sparse data and reducing reliance on large datasets.

Findings

Reduced energy and force errors by up to a factor of two

Enhanced robustness in molecular dynamics simulations

Improved fine-tuning of foundation models with sparse data

Abstract

Machine learning plays an increasingly important role in computational chemistry and materials science, complementing computationally intensive ab initio and first-principles methods. Despite their utility, machine-learning models often lack generalization capability and robustness during atomistic simulations, yielding unphysical energy and force predictions that hinder their real-world applications. We address this challenge by introducing a physics-informed, weakly supervised approach for training machine-learned interatomic potentials (MLIPs). We introduce two novel loss functions, extrapolating the potential energy via a Taylor expansion and using the concept of conservative forces. Our approach improves the accuracy of MLIPs applied to training tasks with sparse training data sets and reduces the need for pre-training computationally demanding models with large data sets. Particularly, we perform extensive experiments demonstrating reduced energy and force errors -- often lower by a factor of two -- for various baseline models and benchmark data sets. Moreover, we demonstrate improved robustness during MD simulations of the MLIP models trained with the proposed weakly supervised loss. Finally, our approach improves the fine-tuning of foundation models on sparse, highly accurate ab initio data. An implementation of our method and scripts for executing experiments are available at https://github.com/nec-research/PICPS-ML4Sci.