$ ⌘K
// constructs / neural_network_potential

Neural Network Interatomic Potential

id: neural_network_potential

A machine learning model trained on DFT data that predicts atomic energies, forces, and stresses from atomic positions and species. Examples include CHGNet, M3GNet, MACE, and SchNet. Enables molecular dynamics and structure relaxation at near-DFT accuracy but orders of magnitude faster. CHGNet uniquely incorporates magnetic moments and charge states.

// usage

Used in 1 pax

↑ positive relationship 14 findings