N. Bansal

Latest

• Can We Gain More from Orthogonality Regularizations in Training Deep Networks?