Xref: utzoo comp.ai:8311 sci.bio:4221 sci.psychology:3954 alt.cyberpunk:5467
Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!cs.utexas.edu!swrinde!mips!apple!portal!cup.portal.com!mmm
From: mmm@cup.portal.com (Mark Robert Thorson)
Newsgroups: comp.ai,sci.bio,sci.psychology,alt.cyberpunk
Subject: Re: The Bandwidth of the Brain
Message-ID: <37351@cup.portal.com>
Date: 29 Dec 90 07:19:54 GMT
References: <37034@cup.portal.com>
Organization: The Portal System (TM)
Lines: 70

Mark Hopkins says:

> In the early stages of development, this neural net can be trained
> (say by backpropagation) so as to minimize the difference between
> first stage input and second stage output, thus training the identity
> function on it.  If backpropagation is used in modelling this
> phenomenon, it will be one of the few instances where
> backpropagation can be used for unsupervised learning.
> 
> The significance of being able to emulate the identity function this
> way is the bottleneck that has to be passed through between the
> first and second stages.  Spontaneous feature discovery/extraction
> is forced on the neural net.
>  
> What you will find in the brain, that corresponds to this, is
> RECURRENCE along the visual pathway.  That is, connections that go
> toward the visual cortex from the eye, AND connections that go
> from the visual cortex in the direction of the eye.
>  
> I bet you that the recurrence is there to perform the function
> described above.

I agree, except I doubt that it's restricted to the early stages of development.
I think it happens continuously.  For example, our
eyes may respond differently after a minute of reading compared
to a minute of driving.  You can see almost the same idea in the
following excerpt from THE THINKING MACHINE by C. Judson Herrick
(Univ. of Chicago Press, 1929).  His "organic tension" may be the force
that demands spontaneous feature discovery/extraction from the
network.

----------------------------------------------------------

It is not generally recognized that the stream of nervous impulses
between sense-organs and brain is not one-way traffic.  Most of
these nervous impulses are directed inward toward the brain, but
there is a respectable amount of transmission in the reverse
direction.

This is most evident in the eye, for the optic nerve contains a very
large number of nerve fibers which conduct from brain to retina.
Just what the effect of these outgoing nervous currents upon the
retina may be is not very clear.  In some fishes they have been
shown to cause changes in the length of rods and cones and in the
retinal pigment, and probably in our own eyes they have some effect
to alter the sensitivity of the retina to light.  Besides these outgoing
fibers in the optic nerve there are fibers in other nerves that activate
the accessory organs of vision, those, for instance, that change the
size of the pupil and the focus of the eye.

Most sense organs are under some sort of central control of this sort,
and inside the brain there are similar nervous connections which
may activate or sensitize the lower sensory centers.

Conscious attention to anything going on outside implies an organic
tension within the nervous system and this tension may extend
outward even to the sense-organ itself so that the sense of sight, for
instance, becomes more acute when we are straining the eyes to
read a distant street-sign.  The eye serves the mind and the mind,
in turn, serves the eye.

The muscles have a sensory nerve supply which is as important as
their motor nerve supply.  The "muscular branch" of a nerve, say to
the biceps, has about one-third as many sensory fibers as motor
fibers, so that every change in the contraction of the muscle is
directly reported back to the brain.  These muscular and similar
"intimate" sensations rarely are clearly recognized, yet they play
a tremendously important part as the organic background of our
conscious attitudes and reactions.  They are essential parts of our
thinking machinery.