The cortical regions in the FARS model can be classified into two categories based on their implementation:
A layer is an array of columns (see previous section "Micro-Schema" ). A layer is updated by first computing the membrane potentials of all the columns and then their activities, thus implementing a parallel update scheme. Major regions in the model, such as F5, AIP, Mcx, SI, SII, F2 are implemented as layers.
Please see layer.h and layer.c for the codes of the layer structure and its member functions.
For computational efficiency, some regions are not implemented as layers; they are simply implemented as sensors. A sensor is an array of values. It is used for distributed representation of parameters, such as the width of an object. Regions like PIP, SMA, F6, Area 46, etc, are implemented as sensors.
Please see sensor.h and sensor.c and for the codes of the sensor structure and its member functions.