Implement the Random Hierarchy Model (RHM) in PyTorch and train (deep) neural networks or kernels on it.
Code for the paper "How Deep Neural Networks Learn Compositional Data: The Random Hierarchy Model", preprint available here: arXiv:2307.02129.
pcsl-epfl/hierarchy-learning
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|