Li–P–S Electrolyte Materials as a Benchmark for Machine-Learned Interatomic Potentials
We introduce LiPS-25, a dataset and testing framework for evaluating the accuracy and physical reliability of MLIPs for crystalline and amorphous Li–P–S solid-state electrolytes.
Atomic cluster expansion potential for the Si-H system
We develop a bespoke general-purpose MLIP for the Si-H binary system, for "off-the-shelf" fast and accurate predictions, paving the way toward realistic simulations of a-Si:H in devices.
Machine-learning-driven modelling of amorphous and polycrystalline BaZrS3
We present a bespoke MLIP model for the emerging solar-cell material BaZrS3, which can be used to simulate both the amorphous precursor and the polycrystalline material with complex grain boundaries.
Accelerated First-Principles Exploration of Structure and Reactivity in Graphene Oxide
We describe the atomic structure of graphene oxide (GO) using ML-driven simulations, comparing our simulations to earlier experiments, and providing the GO-MACE-23 model for future studies in the field.
