Silicon MLIP Datasets
Training, validation, and testing datasets for the publication “Discrepancies and error evaluation metrics for machine learning interatomic potentials.” Each directory contains compressed structural data (VASP POSCAR format) along with the corresponding energies, forces, and distance data used to benchmark MLIP performance on silicon.
Reference: Yunsheng Liu, Xingfeng He, Yifei Mo*, “The Discrepancies and Error Evaluation Metrics for Machine Learning Interatomic Potentials on Simulating Atom Dynamics”, NPJ Computational Materials 9, 174 (2023)