On the Implicit Bias of Initialization Shape: Beyond Infinitesimal Mirror Descent
arXiv:2102.09769
Abstract
Recent work has highlighted the role of initialization scale in determining the structure of the solutions that gradient methods converge to. In particular, it was shown that large initialization leads to the neural tangent kernel regime solution, whereas small initialization leads to so called "rich regimes". However, the initialization structure is richer than the overall scale alone and involves relative magnitudes of different weights and layers in the network. Here we show that these relative scales, which we refer to as initialization shape, play an important role in determining the learned model. We develop a novel technique for deriving the inductive bias of gradient-flow and use it to obtain closed-form implicit regularizers for multiple cases of interest.
arXiv comment33 pages, 2 figures
References in corpus (8)
- Neural Tangent Kernel: Convergence and Generalization in Neural Networks
- On Lazy Training in Differentiable Programming
- Gradient descent aligns the layers of deep linear networks
- Gradient Descent Maximizes the Margin of Homogeneous Neural Networks
- Implicit Regularization in Deep Learning May Not Be Explainable by Norms
- Implicit Regularization for Optimal Sparse Recovery
- On the Proof of Global Convergence of Gradient Descent for Deep ReLU Networks with Linear Widths
- Implicit Regularization in ReLU Networks with the Square Loss