Poster
0

Geodesic Mode Connectivity

Avatar
Voices Powered byElevenlabs logo

Authors

Charlie Tan, Theodore Long, Sarah Zhao, Rudolf Laine

Abstract

Mode connectivity is a phenomenon where trained models are connected by a path of low loss. We reframe this in the context of Information Geometry, where neural networks are studied as spaces of parameterized distributions with curved geometry. We hypothesize that shortest paths in these spaces, known as geodesics, correspond to mode-connecting paths in the loss landscape. We propose an algorithm to approximate geodesics and demonstrate that they achieve mode connectivity.

Follow Us on

0 comments

Add comment
Recommended SciCasts