A Scale-Invariant Diagnostic Approach Towards Understanding Dynamics of Deep Neural Networks

2024/07/12 by Ambarish Moharil, Moharil, Ambarish, Damian A. Tamburri +7
Computer Science · #Artificial Intelligence (cs.AI) #FOS: Computer and information sciences #Neural Networks and Applications #Neural and Evolutionary Computing (cs.NE)

paper · pdf · doi:10.48550/arxiv.2407.09585

openalex publication_date 2024/07/12 · openalex created_date 2024/07/17 · openalex updated_date 2026/07/28

Abstract

This paper introduces a scale-invariant methodology employing Fractal Geometry to analyze and explain the nonlinear dynamics of complex connectionist systems. By leveraging architectural self-similarity in Deep Neural Networks (DNNs), we quantify fractal dimensions and roughness to deeply understand their dynamics and enhance the quality of intrinsic explanations. Our approach integrates principles from Chaos Theory to improve visualizations of fractal evolution and utilizes a Graph-Based Neural Network for reconstructing network topology. This strategy aims at advancing the intrinsic explainability of connectionist Artificial Intelligence (AI) systems.

Related

• Recursive Self-Similarity in Deep Weight Spaces of Neural Architectures: A Fractal and Coarse Geometry Perspective
• FractalNet: Ultra-Deep Neural Networks without Residuals
• Computing brain networks with complex dynamics
• Fractal geometry predicts dynamic differences in structural and functional connectomes
• On The Potential of The Fractal Geometry and The CNNs Ability to Encode it