OMol-25

The OMol25 dataset is a collection of molecular crystal structures used to train machine learning interatomic potentials for evaluating their accuracy in crystal structure prediction.

Papers using OMol-25 (7)

MACE-POLAR-1: A Polarisable Electrostatic Foundation Model for Molecular Chemistry2026 · 3 citesDFT Accuracy on Crystal Structure Prediction with Machine Learning Interatomic Potentials2026Transformers Discover Molecular Structure Without Graph Priors2025 · 2 citesA recipe for scalable attention-based MLIPs: unlocking long-range accuracy with all-to-all node attention2026 · 1 citesHow Accurate Are DFT Forces? Unexpectedly Large Uncertainties in Molecular Datasets2025 · 1 citesBenchmarking foundation potentials against quantum chemistry methods for predicting molecular redox potentials2025 · 1 citesScaling Machine Learning Interatomic Potentials with Mixtures of Experts2026