Uncertainty-biased molecular dynamics for learning uniformly accurate interatomic potentials

Abstract

Efficiently creating a concise but comprehensive data set for training machine-learned interatomic potentials (MLIPs) is an under-explored problem. Active learning, which uses biased or unbiased molecular dynamics (MD) to generate candidate pools, aims to address this objective. Existing biased and unbiased MD-simulation methods, however, are prone to miss either rare events or extrapolative regions -- areas of the configurational space where unreliable predictions are made. This work demonstrates that MD, when biased by the MLIP's energy uncertainty, simultaneously captures extrapolative regions and rare events, which is crucial for developing uniformly accurate MLIPs. Furthermore, exploiting automatic differentiation, we enhance bias-forces-driven MD with the concept of bias stress. We employ calibrated gradient-based uncertainties to yield MLIPs with similar or, sometimes, better accuracy than ensemble-based methods at a lower computational cost. Finally, we apply uncertainty-biased MD to alanine dipeptide and MIL-53(Al), generating MLIPs that represent both configurational spaces more accurately than models trained with conventional MD.

References in corpus (16)

• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• Well-Tempered Metadynamics: A Smoothly Converging and Tunable Free-Energy Method
• ANI-1: An extensible neural network potential with DFT accuracy at force field computational cost
• E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentials
• The Open Catalyst 2020 (OC20) Dataset and Community Challenges
• Sketching as a Tool for Numerical Linear Algebra
• SchNet: A continuous-filter convolutional neural network for modeling quantum interactions
• TorchMD: A deep learning framework for molecular simulations
• Gaussian Moments as Physically Inspired Molecular Descriptors for Accurate and Scalable Machine Learning Potentials
• Differentiable sampling of molecular geometries with uncertainty-based adversarial attacks
• Fast Uncertainty Estimates in Deep Learning Interatomic Potentials
• Deep Ensembles vs. Committees for Uncertainty Estimation in Neural-Network Force Fields: Comparison and Application to Active Learning
• Fast and Sample-Efficient Interatomic Neural Network Potentials for Molecules and Materials Based on Gaussian Moments
• The long road to calibrated prediction uncertainty in computational chemistry
• Metadynamics sampling in atomic environment space for collecting training data for machine learning potentials
• Self-Tuning Hamiltonian Monte Carlo for Accelerated Sampling

Cited by in corpus (11)

• Enhanced Sampling in the Age of Machine Learning: Algorithms and Applications
• ArcaNN: automated enhanced sampling generation of training sets for chemically reactive machine learning interatomic potentials
• Random sampling versus active learning algorithms for machine learning potentials of quantum liquid water
• Prediction rigidities for data-driven chemistry
• Uncertainty Quantification in Atomistic Simulations of Silicon using Interatomic Potentials
• Enhanced sampling of robust molecular datasets with uncertainty-based collective variables
• Efficient ensemble uncertainty estimation in Gaussian Processes Regression
• Free energy profiles for chemical reactions in solution from high-dimensional neural network potentials: The case of the Strecker synthesis
• Adaptive Equilibration of Molecular Dynamics Simulations
• Learning atomic forces from uncertainty-calibrated adversarial attacks
• Gradient-based grand canonical optimization enabled by graph neural networks with fractional atomic existence