On Wed, 05 Sep 2001 09:37:00 GMT, remove_this!helbrecht at gmx.net
>Kalman Rubinson <kr4 at nyu.edu> wrote:
>>>A neurons sums,
>> temporally and spatially, all the signals impinging on it and fires an
>> output signal or signals depending on the inputs and the membrane
>> properties of the neuron.
>>Is that true? A neuron sums the signals impinging? Or does it any
>other operation on it? Maybe multiplicating, processing them as
>vectors or something like that? No, that isn't a joke - just a
>verifying question (see below).
Good question but I did put in the qualifier: "depending on the
inputs and the membrane properties of the neuron." Those properties
may not be uniform throughout the cell permitting the response to be
non-algebraic and complex.
>If that is true - one further question: It should be impossible -
>only looking to the output signal of the outputting neuron (A) -
>to determine which of its input neurons (B) gave A _what_ signal
>or _how_ it "looked like". Right?
Dunno about "looked like" but, depending on the cell, one might or
might not be able to tell from its output which of its inputs were
>Yes, I know that it's impossible to get the summands ot of a given
>sum. Because of that i am asking so detailed, if neurons, in fact,
>just do summing up the incoming signals. Just to verify that i
>didn't miss a point.
Perhaps, rather than sums, I should have used a term that indicates
that the output is the result of the influences of the inputs, their
distribution and the static/dynamic properties of the neuron.