Gradient flow dynamics of shallow ReLU networks for square loss and orthogonal inputs
Boursier, Etienne, Pillaud-Vivien, Loucas, Flammarion, Nicolas
–arXiv.org Artificial Intelligence
The training of neural networks by gradient descent methods is a cornerstone of the deep learning revolution. Yet, despite some recent progress, a complete theory explaining its success is still missing. This article presents, for orthogonal input vectors, a precise description of the gradient flow dynamics of training one-hidden layer ReLU neural networks for the mean squared error at small initialisation. In this setting, despite non-convexity, we show that the gradient flow converges to zero loss and characterise its implicit bias towards minimum variation norm. Furthermore, some interesting phenomena are highlighted: a quantitative description of the initial alignment phenomenon and a proof that the process follows a specific saddle to saddle dynamics.
arXiv.org Artificial Intelligence
Oct-31-2022
- Country:
- North America > United States (0.14)
- Europe > Switzerland (0.04)
- Genre:
- Research Report (0.50)
- Technology: