A processing port and modification of the Leabra neural modeling framework. Based on a python port of Leabra 8.0 (TODO add link).

• 2022-01-28: This is heavily in development; nothing can be expected to work!

1. Intall the required processing libraries: TheMidiBus (for controlling sketches with MIDI)
2. clone the repository, e.g. into your Processing sketchbook folder
3. Open the LeabraLite.pde sketch in Processing

1. Connection class is abstract, and is implemented by
   1. LayerConnection: standard connection between layers, but weights are also connectable using ConnectableWeight class
   2. DendriteConnection: allows connection to a connection; typically used for modulation of a connection in terms of gating or learning rate
   3. ReservoirConnection: connection between layers and Reservoir instances, which can be used to model intercellular space, and non-synaptic connections
2. Support for NetworkModules; this is a java interface which can be implemented and added to a network; typically a collection of Layers and Connections. The thought is that this may help tidy up creating "brain modules" like Amygdala, Basal Ganglia etc

1. Unit:
2. Layer
3. Connection
4. Network

1. Reservoir: set of leaky integrators for modeling intercellular accumulation of neuromodulators

1. ChoiceModule: integrates value for alternatives until a choice is made
2. EffortModule: recruits a neural population based on a control signal, to be directed at a population in need of extra excitation
3. ValenceLearningModule: increases weights of avoid or approach pathways for a set of properties based on positive and negative valence
4. BasalGangliaModule: simple BG module which control "gas", "break" for behaviours. Currently has only on/off behaviours, not graded responses
5. RuleModule: input-output mapping with definable weights which can be used to model rules for experimental tasks; use DendriteConnection to turn on and off rules

• TestTemplate.pde: template for testing networks
• ModuleTemplate.pde: template for modules
• TestModuleTemplate.pde: template for testing modules

TODO

TODO