Cormorant: Covariant Molecular Neural Networks
This addresses the challenge of accurately modeling complex molecular systems for computational chemistry and physics, representing an incremental improvement with a novel method for a known bottleneck.
The authors tackled the problem of learning molecular properties and potential energy surfaces by proposing Cormorant, a rotationally covariant neural network architecture, which significantly outperforms competing algorithms on the MD-17 dataset and is competitive on the GDB-9 dataset.
We propose Cormorant, a rotationally covariant neural network architecture for learning the behavior and properties of complex many-body physical systems. We apply these networks to molecular systems with two goals: learning atomic potential energy surfaces for use in Molecular Dynamics simulations, and learning ground state properties of molecules calculated by Density Functional Theory. Some of the key features of our network are that (a) each neuron explicitly corresponds to a subset of atoms; (b) the activation of each neuron is covariant to rotations, ensuring that overall the network is fully rotationally invariant. Furthermore, the non-linearity in our network is based upon tensor products and the Clebsch-Gordan decomposition, allowing the network to operate entirely in Fourier space. Cormorant significantly outperforms competing algorithms in learning molecular Potential Energy Surfaces from conformational geometries in the MD-17 dataset, and is competitive with other methods at learning geometric, energetic, electronic, and thermodynamic properties of molecules on the GDB-9 dataset.