Basic Neuron Questions

KP-PC k.p.collins at worldnet.att.net%remove%
Wed Apr 9 10:02:06 EST 2003


"Didier A. Depireux" <didier at rai.isr.umd.edu> wrote in message
news:b717sa$s6b$1 at grapevine.wam.umd.edu...
| John H. <johnh at faraway.xxx> wrote:
| > At what time scale does the nervous system operate?
|
| Well,  that's still an open question isn't it? There are 2 schools
of
| thought (grossly oversimplifying), one that says timing is
exquisitely
| important, and the relative timing of 2 spikes might encode
information down
| to the 100 microsecond range, and the other that says that only the
average
| firing rate, averaged over a certain time window and over
neighboring
| spikes, is what matters.
| If you look at the book "Spikes" by Bialek and others, they take
the
| position that it's half-way between the time code and the rate
code, i.e.
| you have about 1 spike per relevant time-scale. Say you listen to
speech,
| and extract the spectral envelope which changes (phonemic rate)
_about_ every
| 30 msec. If the primary auditory cortical neurons are the ones in
charge of
| extracting spectral envelope, you will find as a correlate that
primary
| auditory neurons fires at about a rate of 30 Hz...
|
| So the answer, as ever, is "it depends".
|
| For timing issues, look at the paper on
| http://homepage.mac.com/cariani/CarianiWebsite/Publications.html
| called "As if time really mattered"

Thanks for this ref, Didier.

~Quoting from the article:

"ratio codes and central spectra" [Is there an echo in here?]

"reverberation times and oscillations" are artifacts of TD
E/I-minimization.

The 'ratio' does not "code" information, nor do "central spectra",
nor does "reverberation" nor does "oscillation".

All these things do, of course, occur within neural tissue, but it's
Topologically-Distributed sum(force * distance) - =work= that's
rigorously correlated to the neural Topology that encodes
information-content.

The TD E/I-minimization ratio =maps= the directions in which work is
being performed in terms of 'moving toward' and 'moving away from'
WDB2T correlates in the external environment.

That is, the TD E/I-minimization ratio maps the usefulness inherent
in the work inherent in the nerual energydynamics.

Work =is= the information-content.

TD E/I(up), 'randomness'-up = useful work down.

TD E/I(down), 'randomness'-down = useful work up.

In 'striving' to achieve TD E/I-minimization, nervous systems perform
work to imbue themselves with 'usefulness' with respect to 'climbing'
the WDB2T energy-gradient.

'timing' is just one way of addressing the degree of order within the
work, which is one beating-around-the-bush way of addressing degree
of 'usefulness' inherent in the work that nervous systems are
'momentarily' doing with respect to WDB2T.

But why beat around the bush through invocation of stuff that has no
existence within physical reality ['time']?

Why not just see the work that's, through the rigorously-mapped
neural Topology is rigorously in accord with WDB2T ^ -1?

Why not just see the degree of order within the work?

Why not see that  relative TD E/I is the degree of order within the
work?

The Topologically-Distributed work =is= the information-content.

Why not see it?

[Forgive me, please, what's here should be much better stated. I'm
'tired' and hungry, and my hands are freezing. Use your own good
noggin'-labs to find your own ways through it. If I can find a way to
'stoke the internal fire', I'll say it better.]

Cheers, Didier, ken

| Didier
|
| --
| Didier A Depireux         ddepi001 at umaryland.edu
didier at isr.umd.edu
| 685 W.Baltimore Str
http://neurobiology.umaryland.edu/depireux.htm
| Anatomy and Neurobiology                      Phone: 410-706-1272
(off)
| University of Maryland                                      -1273
(lab)
| Baltimore MD 21201 USA                             Fax:
1-410-706-2512





More information about the Neur-sci mailing list