Using Multiple Vector Channels Improves E(n)-Equivariant Graph Neural Networks
This incremental improvement may be of practical use to researchers in machine learning for the physical sciences.
The authors tackled the problem of improving E(n)-equivariant graph neural networks by extending them to use multiple equivariant vectors per node, resulting in enhanced performance across physical system benchmarks like N-body dynamics and molecular property predictions with minimal runtime or parameter increases.
We present a natural extension to E(n)-equivariant graph neural networks that uses multiple equivariant vectors per node. We formulate the extension and show that it improves performance across different physical systems benchmark tasks, with minimal differences in runtime or number of parameters. The proposed multichannel EGNN outperforms the standard singlechannel EGNN on N-body charged particle dynamics, molecular property predictions, and predicting the trajectories of solar system bodies. Given the additional benefits and minimal additional cost of multi-channel EGNN, we suggest that this extension may be of practical use to researchers working in machine learning for the physical sciences