Skip to yearly menu bar Skip to main content


Explicit Regularization in Overparametrized Models via Noise Injection

Antonio Orvieto · Anant Raj · Hans Kersting · Francis Bach

Auditorium 1 Foyer 119


Injecting noise within gradient descent has several desirable features, such as smoothing and regularizing properties. In this paper, we investigate the effects of injecting noise before computing a gradient step. We demonstrate that small perturbations can induce explicit regularization for simple models based on the L1-norm, group L1-norms, or nuclear norms. However, when applied to overparametrized neural networks with large widths, we show that the same perturbations can cause variance explosion. To overcome this, we propose using independent layer-wise perturbations, which provably allow for explicit regularization without variance explosion. Our empirical results show that these small perturbations lead to improved generalization performance compared to vanilla gradient descent.

Live content is unavailable. Log in and register to view live content