Resolving the Body-Order Paradox of Machine Learning Interatomic Potentials

Abstract

In many cases, the predictions of machine learning interatomic potentials (MLIPs) can be interpreted as a sum of body-ordered contributions, which is explicit when the model is directly built on neighbor density correlation descriptors, and implicit when the model captures the correlations through non-linear functions of low body-order terms. In both cases, the "effective body-orderedness" of MLIPs remains largely unexplained: how do the models decompose the total energy into body-ordered contributions, and how does their body-orderedness affect the accuracy and learning behavior? In answering these questions, we first discuss the complexities in imposing the many-body expansion on ab initio calculations at the atomic limit. Next, we train a curated set of MLIPs on datasets of hydrogen clusters and reveal the inherent tendency of the ML models to deduce their own, effective body-order trends, which are dependent on the model type and dataset makeup. Finally, we present different trends in the convergence of the body-orders and generalizability of the models, providing useful insights for the development of future MLIPs.

References in corpus (20)

• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• On representing chemical environments
• Moment Tensor Potentials: a class of systematically improvable interatomic potentials
• Recent developments in the PySCF program package
• Nearsightedness of Electronic Matter
• Physics-inspired structural representations for molecules and materials
• Materials Cloud, a platform for open computational science
• A spin-adapted Density Matrix Renormalization Group algorithm for quantum chemistry
• Scalable Parallel Algorithm for Graph Neural Network Interatomic Potentials in Molecular Dynamics Simulations
• On the Completeness of Atomic Structure Representations
• Evaluation of the MACE Force Field Architecture: from Medicinal Chemistry to Materials Science
• Evidence for supercritical behavior of high-pressure liquid hydrogen
• Feature Optimization for Atomistic Machine Learning Yields A Data-Driven Construction of the Periodic Table of the Elements
• Block2: a comprehensive open source framework to develop and apply state-of-the-art DMRG algorithms in electronic structure and beyond
• Combining DFT and Many-Body Methods to Understand Correlated Materials
• Ground-state properties of the hydrogen chain: insulator-to-metal transition, dimerization, and magnetic phases
• The convergence of the ab-initio many-body expansion for the cohesive energy of solid mercury
• Prediction rigidities for data-driven chemistry
• Rigorous body-order approximations of an electronic structure potential energy landscape
• Convergence of Body-Orders in Linear Atomic Cluster Expansions