Modules

The modules package contains the building blocks for constructing retinal models in OpenRetina. These modular components can be combined to create custom neural network architectures.

Overview

The modules are organized into four main categories:

TODO partly generated by AI, needs re-organising. - Core Modules: Convolutional feature extractors that process spatio-temporal visual input - Readout Modules: Components that map core features to individual neuron responses
- Layers: Custom neural network layers and utilities - Losses: Loss functions specialized for neural response prediction