How Graph Neural Network Interatomic Potentials Extrapolate: Role of the Message-Passing Algorithm

Abstract

Graph neural network interatomic potentials (GNN-IPs) are gaining significant attention due to their capability of learning from large datasets. Specifically, universal interatomic potentials based on GNN, usually trained with crystalline geometries, often exhibit remarkable extrapolative behavior towards untrained domains, such as surfaces or amorphous configurations. However, the origin of this extrapolation capability is not well understood. This work provides a theoretical explanation of how GNN-IPs extrapolate to untrained geometries. First, we demonstrate that GNN-IPs can capture non-local electrostatic interactions through the message-passing algorithm, as evidenced by tests on toy models and DFT data. We find that GNN-IP models, SevenNet and MACE, accurately predict electrostatic forces in untrained domains, indicating that they have learned the exact functional form of the Coulomb interaction. Based on these results, we suggest that the ability to learn non-local electrostatic interactions, coupled with the embedding nature of GNN-IPs, explains their extrapolation ability. We find that the universal GNN-IP, SevenNet-0, effectively infers non-local Coulomb interactions in untrained domains but fails to extrapolate the non-local forces arising from the kinetic term, which supports the suggested theory. Finally, we address the impact of hyperparameters on the extrapolation performance of universal potentials, such as SevenNet-0 and MACE-MP-0, and discuss the limitations of the extrapolation capabilities.

References in corpus (19)

• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentials
• A Universal Graph Deep Learning Interatomic Potential for the Periodic Table
• DP-GEN: A concurrent learning platform for the generation of reliable deep learning based potential energy models
• On the Number of Linear Regions of Deep Neural Networks
• DeePMD-kit v2: A software package for Deep Potential models
• A Fourth-Generation High-Dimensional Neural Network Potential with Accurate Electrostatics Including Non-local Charge Transfer
• Towards Universal Neural Network Potential for Material Discovery Applicable to Arbitrary Combination of 45 Elements
• SpookyNet: Learning Force Fields with Electronic Degrees of Freedom and Nonlocal Effects
• Interatomic potentials for ionic systems with density functional accuracy based on charge densities obtained by a neural network
• MACE: Higher Order Equivariant Message Passing Neural Networks for Fast and Accurate Force Fields
• Scalable Parallel Algorithm for Graph Neural Network Interatomic Potentials in Molecular Dynamics Simulations
• Self-Consistent Determination of Long-Range Electrostatics in Neural Network Potentials
• MatterSim: A Deep Learning Atomistic Model Across Elements, Temperatures and Pressures
• Atomic energy mapping of neural network potential
• Tensor-reduced atomic density representations
• Accelerated identification of equilibrium structures of multicomponent inorganic crystals using machine learning potentials
• Multilayer atomic cluster expansion for semi-local interactions
• How Graph Neural Network Interatomic Potentials Extrapolate: Role of the Message-Passing Algorithm

Cited by in corpus (2)

• Universal machine learning interatomic potentials poised to supplant DFT in modeling general defects in metals and random alloys
• How Graph Neural Network Interatomic Potentials Extrapolate: Role of the Message-Passing Algorithm