Better Neural Network Expressivity: Subdividing the Simplex

Re-assign community