Atomic Foundation Models

3.00 points · University catalogue

Machine-learning and foundation-model approaches for atomistic simulations of materials. Equivariant message-passing neural networks for interatomic forces and for electronic, dielectric, and magnetic degrees of freedom. Heat and charge transport in solids ranging from crystals to disordered and amorphous materials. Hands-on assignments: simulation workflows, data analysis, training, and fine-tuning of various machine-learning models for atomistic simulations of materials