Open Catalyst Project

The Open Catalyst Project is a dataset used to evaluate graph neural networks for predicting relaxed energy in molecular property calculations, facilitating material discovery.

Papers using Open Catalyst Project (4)

Adapting OC20-trained EquiformerV2 Models for High-Entropy Materials2024 · 31 citesAutomatic graph representation algorithm for heterogeneous catalysis2023 · 7 citesImproved Uncertainty Estimation of Graph Neural Network Potentials Using Engineered Latent Space Distances2024 · 6 citesLightweight Geometric Deep Learning for Molecular Modelling in Catalyst Discovery2024