Open Catalyst

The 'Open Catalyst' dataset is a benchmark that contains a large collection of atomic structures and their corresponding potential energy surfaces, used to evaluate the performance of SE(3)-equivariant graph neural networks in 3D atomistic modeling.

Papers using Open Catalyst (14)

A recipe for scalable attention-based MLIPs: unlocking long-range accuracy with all-to-all node attention2026 · 2 citesUncertainty Quantification in Graph Neural Networks with Shallow Ensembles2025 · 2 citesEquiformerV3: Scaling Efficient, Expressive, and General SE(3)-Equivariant Graph Attention Transformers2026Learning 3D Anisotropic Noise Distributions Improves Molecular Force Field Modeling2025Tensor Decomposition Networks for Fast Machine Learning Interatomic Potential Computations2025Beyond Force Metrics: Pre-Training MLFFs for Stable MD Simulations2025Equivariant Spherical Transformer for Efficient Molecular Modeling2025From Molecules to Materials: Pre-training Large Generalizable Models for Atomic Property Prediction2023 · 22 citesFAENet: Frame Averaging Equivariant GNN for Materials Modeling2023 · 7 citesEwald-based Long-Range Message Passing for Molecular Graphs2023 · 6 citesMatSciML: A Broad, Multi-Task Benchmark for Solid-State Materials Modeling2023 · 6 citesTransfer learning for atomistic simulations using GNNs and kernel mean embeddings2023 · 5 citesMolecular Geometry-aware Transformer for accurate 3D Atomic System modeling2023 · 3 citesOn the importance of catalyst-adsorbate 3D interactions for relaxed energy predictions2023