Fetching the paper…

Learning deep linear neural networks: Riemannian gradient flows and convergence to global minimizers · Around