Superpolynomial Lower Bounds for Learning One-Layer Neural Networks using Gradient Descent

Open in new window