Discrepancies and error evaluation metrics for machine learning interatomic potentials
Authors:
Open Access
- 26 September 2023
- journal article
- research article
- Published by Springer Nature in npj Computational Materials
- Vol. 9 (1) , 1-13
Abstract
Machine learning interatomic potentials (MLIPs) are a promising technique for atomic modeling. While small errors are widely reported for MLIPs, an open concern is whether MLIPs can accurately reproduce atomistic dynamics and related physical properties in molecular dynamics (MD) simulations. In this study, we examine the state-of-the-art MLIPs and uncover several discrepancies related to atom dynamics, defects, and rare events (REs), compared to ab initio methods. We find that low averaged errors by current MLIP testing are insufficient, and develop quantitative metrics that better indicate the accurate prediction of atomic dynamics by MLIPs. The MLIPs optimized by the RE-based evaluation metrics are demonstrated to have improved prediction in multiple properties. The identified errors, the evaluation metrics, and the proposed process of developing such metrics are general to MLIPs, thus providing valuable guidance for future testing and improvements of accurate and reliable MLIPs for atomistic modeling.Keywords
All Related Versions
Funding Information
- National Science Foundation (2004837, 1940166)
This publication has 45 references indexed in Scilit:
- Python Materials Genomics (pymatgen): A robust, open-source python library for materials analysisComputational Materials Science, 2013
- A high-throughput infrastructure for density functional theory calculationsComputational Materials Science, 2011
- Atom-centered symmetry functions for constructing high-dimensional neural network potentialsThe Journal of Chemical Physics, 2011
- Gaussian Approximation Potentials: The Accuracy of Quantum Mechanics, without the ElectronsPhysical Review Letters, 2010
- Generalized Neural-Network Representation of High-Dimensional Potential-Energy SurfacesPhysical Review Letters, 2007
- Calculations of Silicon Self-Interstitial DefectsPhysical Review Letters, 1999
- Intrinsic point defects in crystalline silicon: Tight-binding molecular dynamics studiesof self-diffusion, interstitial-vacancy recombination, and formation volumesPhysical Review B, 1997
- Rationale for mixing exact exchange with density functional approximationsThe Journal of Chemical Physics, 1996
- Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis setPhysical Review B, 1996
- Atomistic calculation of the self-interstitial diffusivity in siliconApplied Physics Letters, 1993