Skip to yearly menu bar Skip to main content


Poster

An SDE for Modeling SAM: Theory and Insights

Enea Monzio Compagnoni · Luca Biggio · Antonio Orvieto · Frank Proske · Hans Kersting · Aurelien Lucchi

Exhibit Hall 1 #204
[ ]
[ Slides [ PDF [ Poster

Abstract:

We study the SAM (Sharpness-Aware Minimization) optimizer which has recently attracted a lot of interest due to its increased performance over more classical variants of stochastic gradient descent. Our main contribution is the derivation of continuous-time models (in the form of SDEs) for SAM and two of its variants, both for the full-batch and mini-batch settings. We demonstrate that these SDEs are rigorous approximations of the real discrete-time algorithms (in a weak sense, scaling linearly with the learning rate). Using these models, we then offer an explanation of why SAM prefers flat minima over sharp ones -- by showing that it minimizes an implicitly regularized loss with a Hessian-dependent noise structure. Finally, we prove that SAM is attracted to saddle points under some realistic conditions. Our theoretical results are supported by detailed experiments.

Chat is not available.