Current neural networks are incapable of understanding symmetry, which limits the conclusions that can be drawn from the data. Researchers from Lehigh and Stanford universities developed a new model that includes symmetry-aware features to improve symmetry approximation.
Read More