Generalizing Denoising to Non-Equilibrium Structures Improves Equivariant Force Fields

arXiv:2403.09549v1 Announce Type: new
Abstract: Understanding the interactions of atoms such as forces in 3D atomistic systems is fundamental to many applications like molecular dynamics and catalyst design. However, simulating these interactions requires compute-intensive ab initio calculations and thus results in limited data for training neural networks. In this paper, we propose to use denoising non-equilibrium structures (DeNS) as an auxiliary task to better leverage training data and improve performance. For training with DeNS, we first corrupt a 3D structure …

abstract applications arxiv catalyst compute cs.ai cs.lg data denoising design dynamics equilibrium fields however interactions molecular dynamics networks neural networks paper physics.comp-ph results systems training type understanding