Our machine-learned interatomic potentials (MLIPs)
SiOx-ACE-24
An ACE potential for the full Si–O system capable of studying very-high-pressure silica, surfaces and aerogels, and amorphous phases beyond few-nanometre length scales.
Nature Communications, 2024 | Zenodo
GO-MACE-23
A MACE potential for thermal reduction of graphene oxide.
Angewandte Chemie International Edition, 2024 | Zenodo
GST-GAP-22
A Gaussian approximation potential model for thermal-induced phase change of germanium–antimony–tellurium compositions up to device-scale.
Nature Electronics, 2023 | Zenodo
SiO2-GAP-22
A Gaussian approximation potential model for the thermodynamic properties of crystalline and amorphous single-phase bulk silica.
npj Computational Materials, 2022 | Zenodo
Our datasets
a-Si-24
A dataset of amorphous silicon structures, each representing the final snapshot of a distinct melt-quench trajectory, to investigate paracrystallinity in amorphous silicon.
arXiv preprint
a-Si-23
A dataset of million-atom-scale amorphous silicon structures from MD trajectories driven by a potential generated using the teacher-student approach for structural defect analysis.
Angewandte Chemie International Edition, 2024 | Zenodo
C-SYNTH-23M
A synthetic dataset of carbon structures from MD trajectories driven by the C-GAP-17 potential model.
Digital Discovery, 2023 | Github | Zenodo
