Artificial Consciousness/Neural Correlates/Neural Network Models/Instar Model

From Wikiversity
Jump to navigation Jump to search

The Instar Model[edit | edit source]

To simplify the input model, so that we don't have to take into account dendritic complexity, Theorists started out by assuming that every input was equivalent to the next except for the information flowing down it. The simplest model of the input network would then, be something called an instar, where all connections were the same length, and of equivalent electrical characteristics, so that all that was important was the signal flowing down it.

Of course this is a theoretical absurdity, really only good for stylizing the nature of input. Such a mechanism, it turned out to be was a gatherer of signals, in essence a form of Multiplexor that took separate signals and produced a single complex aggregation of signals, as an output to the actual cell.

It was then thought that if we could translate dendritic connections into a mathematical formula, we would find equivalence between different individuals at the mathematical level. The neuron was seen as a simultaneous equation expressed by the dendrites, that fed a processing section in the Neuron and the output of the neuron would be the result of that processing. Problem was there was no mathematical equivalency to be found in the so called simultaneous equations.

Worse from the point of view of predicting the outputs of the cell, different signals took different times to travel through the dendrites and arrive at the cell, because the nature of the processes that they traveled down were inimical to electrical current. As a result, the original instar Model is only used today as an abstraction.