GEOM

The GEOM dataset provides rich 3D conformational diversity of molecular structures and is used to evaluate the effectiveness of self-supervised learning methods for molecular representation.

Papers using GEOM (6)

UniGenX: a unified generative foundation model that couples sequence, structure and function to accelerate scientific design across proteins, molecules and materials2025 · 5 citesStiefel Flow Matching for Moment-Constrained Structure Elucidation2024 · 2 citesLearning the Neighborhood: Contrast-Free Multimodal Self-Supervised Molecular Graph Pretraining2025EquiBoost: An Equivariant Boosting Approach to Molecular Conformation Generation2025Riemannian Denoising Model for Molecular Structure Optimization with Chemical Accuracy2024Reflection-Equivariant Diffusion for 3D Structure Determination from Isotopologue Rotational Spectra in Natural Abundance2023