Depth Separations in Neural Networks: Separating the Dimension from the Accuracy

Re-assign community