Fetching the paper…

Existence, uniqueness, and convergence rates for gradient flows in the training of artificial neural networks with ReLU activation · Around