Probing the Latent Hierarchical Structure of Data via Diffusion Models

• Main Authors: Sclocchi, Antonio, Favero, Alessandro, Levi, Noam Itzhak, Wyart, Matthieu
• Format: Journal Article
• Language: English
• Published: 17.10.2024
• Subjects: Computer Science - Learning; Physics - Disordered Systems and Neural Networks; Statistics - Machine Learning

High-dimensional data must be highly structured to be learnable. Although the compositional and hierarchical nature of data is often put forward to explain learnability, quantitative measurements establishing these properties are scarce. Likewise, accessing the latent variables underlying such a dat...