- A set of tools for analysing deep fully-connected recitifer networks i.e. networks with ReLU activation.
- Specific tools include methods for analysing optimisation, variable disentangling, variable size representation, sparsity, and generalisation.
- The ultimate goal is to develop useful mathematical models for deep rectifier networks, a model for supervised learning that is not yet well-understood.
This repository is still in alpha mode and not yet pip installable.