Deep network as memory space: complexity, generalization, disentangled representation and interpretability

TL;DR

This paper proposes a geometrical framework linking deep networks to physics principles, viewing them as memory spaces where complexity and generalization are interconnected, enhancing interpretability.

Contribution

It introduces a novel geometrization approach applying physics concepts to analyze deep networks, focusing on complexity, generalization, and disentangled representations.

Findings

Fisher metric formulation of network complexity

Application of least action principle to deep networks

Geometry of deep network configurations

Abstract

By bridging deep networks and physics, the programme of geometrization of deep networks was proposed as a framework for the interpretability of deep learning systems. Following this programme we can apply two key ideas of physics, the geometrization of physics and the least action principle, on deep networks and deliver a new picture of deep networks: deep networks as memory space of information, where the capacity, robustness and efficiency of the memory are closely related with the complexity, generalization and disentanglement of deep networks. The key components of this understanding include:(1) a Fisher metric based formulation of the network complexity; (2)the least action (complexity=action) principle on deep networks and (3)the geometry built on deep network configurations. We will show how this picture will bring us a new understanding of the interpretability of deep learning systems.