Jump to content

Artificial Consciousness/Neural Correlates/Synaptic Models

From Wikiversity

Synaptic Models

When Hebb wanted to research the nature of the Nerve, he had a problem, Ideally he wanted to explain human nerves, but that was impractical, so he made a decision to try to explain the nerves of Giant Squids instead. What this allowed him to do was to place relatively large by today's standards contacts into the nerve to discover what was happening electrically to the nerve. What he found was a tenuous connection between nerves, that indicated that there was interconnections between the nerves, that was dependent on the signal strength of the source nerve in the response strength of the receiving nerve.

In order to explain this connection, he postulated the existence of something called a "Synapse" His simple mathematical model of synapse function, made the nervous system seem less mysterious, and is the basis for much of our current understanding of how neurons function. The basic Hebb Synapse is expressed mathematically as follows:

Where I is the output value from the source neuron

and W is a Weight determined by the activity of the Synapse


Scientists have since determined that most synapses are chemical in transmission, and electrical in effect. This has been traced down to specific mechanisms in the cell membrane of the Nerve Cell. Exceptions have been found, where direct electrical connection is made across the synapse but these short circuit the synapse and are much less useful, since they can't learn by adjusting the weights, as the chemical/electric synapses can.

Of interest to any synaptic model of the future, will be the new science that is working down into the neurochemistry of the cell, to find out how and why the synapses work the way they do, and what connection there is between the synapse and the growth of neural fibrils such as dendrites and axons. The following models are transient expressions of today's understanding of synapse function