Copy Link
Add to Bookmark
Report
Neuron Digest Volume 05 Number 50
Neuron Digest Friday, 1 Dec 1989 Volume 5 : Issue 50
Today's Topics:
summary : connectinist simulators
Re: Neural net simulators in Public domain?
Re: Good primer on neural-nets?
NN for the Macintosh
Re: NN for the Macintosh
bp simulator executables
Re: bp simulator executables
Simulation program for the Neocognitron. Does it exist ?
Wanted N Nsource + suggestions
Looking for simulator
"Neuralsource" on disk?
Test Data for NN
SN2
Neural Network Simulator Available
neural net software wanted
GMU bps update
new book announcement
Neural Networks on Transputers
Connectionist models of music
musical applications of neural nets
Re: MUSIC and Neural Nets
connectionist models of music
Re: MUSIC and Neural Nets
Send submissions, questions, address maintenance and requests for old issues to
"neuron-request@hplabs.hp.com" or "{any backbone,uunet}!hplabs!neuron-request"
Use "ftp" to get old issues from hplpm.hpl.hp.com (15.255.176.205).
------------------------------------------------------------
Subject: summary : connectinist simulators
From: knareddy@umn-d-ub.D.UMN.EDU (krishna nareddy)
Organization: U of Minnesota-Duluth, Information Services
Date: 29 Sep 89 15:27:59 +0000
In response to the many requests to summarize the availability of
connectionist simulators, here is the list of connectionist simulators I
know about:
(1) The McClelland & Rumelhart book
``Experiments in Parallel Distributed Processing: Handbook of Programs
Experiments and Models'',
comes with a disk of public domain neural network software. Go for it!
(2) MIRRORS/II Connectionist Simulator Available
MIRRORS/II is a general-purpose connectionist simulator
which can be used to implement a broad spectrum of connec-
tionist (neural network) models.
MIRRORS/II is implemented in Franz Lisp and will run
under Opuses 38, 42, and 43 of Franz Lisp on UNIX systems.
It is currently running on a MicroVAX, VAX and SUN 3. It c
an be obtained at no charge via tape or ftp.
If you are interested in obtaining a copy of the software
send your U.S. Mail address via e-mail to
mirrors@cs.umd.edu
or
...!uunet!mimsy!mirrors
or send your U.S. Mail address to
Lynne D'Autrechy
University of Maryland
Department of Computer Science
College Park, MD 20742
and they will send you back a license which you must sign and
return to us and further instructions on how to obtain the
MIRRORS/II software and manual.
(3) There is a connectionist simulator available with the University of
Rochester. It can be ftp'd from cs.rochester.edu free of cost. It comes
with a graphics interface for Sun (Sunview) and can also run without a
graphic i/f. BTW, it is made for the SUN workstation.
It is available for free by anonymous ftp in pub/rcs at cs.rochester.edu.
If you cannot ftp, send a mail request to:
Department of Computer Science
University of Rochester
Rochester, NY 14627
c/o Ms. Peg Meeker
for a tape and manual at about $150/-
(4) There is a package for Mac II developed at Univ. of Colorado, Boulder.
It is called Mactivation and can be ftp'd from pub dir of
boulder.colorado.edu.
(5) George Mason Univ. has a back prop simulator
We have a back prop simulator that is user-friendly and robust, comes with
a user tutorial, incorporates some new speedup techniques, and runs on
8088, 802/386 processors unMS-DOS as well as under Unix (actually DEC
Ultrix). I can make one or more flavors of the object code available to
users via ftp if desired; the source is very portable C code, needing at
most some modification as regards calling a random number generator; only
two lines were changed in porting the MS_Dos version to a VAX 8530 under
Ultrix. Source code can be licensed from my University, but pricing, etc.
hasn't been work out as yet.
Eugene M. Norris
Associate Professor of Computer Science
George Mason University
Fairfax, VA 22030 (703)323-2713 enorris@gmuvax2.gmu.edu
Please contact me if I can be of any further help.
Thank you,
krishna
knareddy@ub.d.umn.edu
(218)726-7664 (w)
(218)728-6551 (h)
------------------------------
Subject: Re: Neural net simulators in Public domain?
From: jochenru@cosmo.UUCP (Jochen Ruhland)
Organization: CosmoNet, D-3000 Hannover 1, FRG
Date: 04 Oct 89 04:05:19 +0000
[[ Editor's Note: I cleaned up spelling though not syntax for our German
colleague. -PM ]]
You're looking on simulators?
We got one ...
It's a PC based Program for backpropagation networks, not ordered in
layers, only loop-free networks required.
Training is done with several numerical methods, including higher nonlinear
methods.
Program includes network layout with mouse input, documented training,
individual input patters for single step simulation (sounds difficult)
Program requires mouse, EGA or VGA card; we haven't tried it on other displays.
If you're interested, send a self-addressed envelope and a floppy disk 5.25
or 3.5" to:
Universitt Kassel
Fachbereich Mathematik
Forschungsgruppe Neuronale Netzwerke
Heinrich-Plett-Str 40
3500 Kassel
West Germany
------------------------------
Subject: Re: Good primer on neural-nets?
From: dg1v+@andrew.cmu.edu (David Greene)
Organization: Graduate School of Industrial Administration, Carnegie Mellon, Pittsburgh, PA
Date: 05 Oct 89 12:24:24 +0000
> Excerpts from netnews.comp.ai.neural-nets: 29-Sep-89 Good primer on
> neural-nets? Steve Wampler@naucse.UUC (285)
> Would some one recommend a good primer on the new approaches on
> neural-nets? (The last I've seen was a *very* old book by M. Minsky
> that was pretty low-level - I'd like something more up to date.)
I'm not sure how detailed an intro you want, but I found:
"The Neural Networks Primer" -- Maureen Caudill for AI Expert magazine
to be a quick and simple introduction. This is actually a series of 8
installments reprinted from AI Expert magazine. The articles introduce
ideas such as: backprop, delta rule, Kohonen learning, and Adaptive
Resonance Theory...
I'd be interested in getting other peoples opinions on the utility of
this "Primer"...
- -David
- --------------------------------------------------------------------
David Perry Greene || ARPA: dg1v@andrew.cmu.edu
GSIA /Robotics || dpg@isl1.ri.cmu.edu
Carnegie Mellon Univ. || BITNET: dg1v%andrew@vb.cc.cmu.edu
Pittsburgh, PA 15213 || UUCP: !harvard!andrew.cmu.edu!dg1v
- --------------------------------------------------------------------
"You're welcome to use my opinions, just don't get them all wrinkled."
------------------------------
Subject: NN for the Macintosh
From: osu-20.ircc.ohio-state.edu!chovan-j@tut.cis.ohio-state.edu (John Chovan)
Organization: The Ohio State University, Columbus, Ohio
Date: 12 Oct 89 18:42:31 +0000
Does anyone have any literature/recommendations for neural network similation
software for a Macintosh SE? Any help would be greatly appreciated.
John Chovan
Chovan-J@OSU-20.IRCC.OHIO-STATE.EDU
[[ Editor's Note: U. Colorado's Mactivation and Rumelhart & McLelland's PDP
Vol III are both for the Mac, and at low/no cost. There are also a number
of commerical ($$$$) packages. The Rochester Simulator has an unsupprted
port of its 4.1 version for the Mac, though RCS is now at 4.2. See otehr
messages in this Digest. -PM ]]
------------------------------
Subject: Re: NN for the Macintosh
From: bwk@mbunix.mitre.org (Barry W. Kort)
Organization: The MITRE Corporation, Bedford, Mass.
Date: 20 Oct 89 18:16:41 +0000
In article <12533571132036@osu-20.ircc.ohio-state.edu>
CHOVAN-J@osu-20.ircc.ohio-state.edu (John Chovan) writes:
> Does anyone have any literature/recommendations for neural network
> simulation software for a Macintosh SE? Any help would be greatly
> appreciated.
Try Mactivation 2.3, Mike Kranzdorf, author:
University of Colorado
Optoelectronic Computing Systems Center
Campus Box 525
Boulder, Colorado 80309 - 0525
(303) 492-8099
mikek@boulder.colorado.EDU
{hao | nbires}!boulder!mikek
Future versions of Mactivation will be available from:
Oblio, Inc.
5943 Sugarloaf Road
Boulder, Colorado 80309
(303) 444-7179
------------------------------
Subject: bp simulator executables
From: enorris@gmuvax2.gmu.edu (Gene Norris)
Organization: George Mason University, Fairfax Va.
Date: 23 Oct 89 21:44:25 +0000
The George Mason University back prop simulator is now available via
anonymous ftp from anonymous@gmuvax2.gmu.edu. After logging on, cd to /nn
and ftp *. The compressed and tar'ed file contains a UNIX executable, an
8088/MS-DOS and an 80286/80386/DOS executable, along with some sample files
for training XOR nets and a user tutorial. The tutorial is in microsoft
Word 4.0 format for a Macintosh and in an unformatted ASCII file with line
breaks, but no diagrams.
(remember to ftp as binary)
------------------------------
Subject: Re: bp simulator executables
From: enorris@gmuvax2.gmu.edu (Gene Norris)
Organization: George Mason Univ. Fairfax, Va.
Date: 24 Oct 89 14:11:02 +0000
In article <512@gmuvax2.gmu.edu> I wrote, somewhat erroneously:
>The George Mason University back prop simulator is now available
>via anonymous ftp from anonymous@gmuvax2.gmu.edu.
The files are available as advertised after you cd /norris. Sometime soon
this directory name will change to /nn. Sorry for the confusion.
Eugene Norris
CS Dept GMU Fairfax, VA 22032 (703)323-2713
enorris@gmuvax2.gmu.edu
------------------------------
Subject: Simulation program for the Neocognitron. Does it exist ?
From: motcid!demetrio@uunet.uu.net (Demetrakis Demetriou)
Organization: Motorola Inc. - Cellular Infrastructure Div., Arlington Heights, IL 60004
Date: 27 Oct 89 20:43:21 +0000
Name: Demetrakis (James) Demetriou
Phone #: (312) 632-5348
I am trying to simulate the Neocognitron from the paper of Kunihiko
Fukushima
"A Neural Network for Visual Pattern Recognition", IEEE, COMPUTER
magazine,March 1988.
Has anyone done a simulation program for the Neocognitron or
something similar.
The purpose for this simulation is to verify that the Neocognitron
model works and then I am going to implement it with digital hardware,
instead of analog that has been done before.
I have more papers from K. Fukushima if anyone is interested.
Feel free to contact me at any time.
Thank you
------------------------------
Subject: Wanted N Nsource + suggestions
From: frobozz@well.UUCP (Jordan Bortz)
Date: 06 Nov 89 23:09:44 +0000
Hi...I'd like to do a NN sim. in Smalltalk-80, and was wondering if someone
had the requesite C code to get started.... What about articles or on-line
examples??
Jordan
***********************************************************************
* Jordan A. Bortz, Higher Level Software, Santa Cruz, CA *
* well!frobozz frobozz@well.sf.ca.us 408 - 476 - 8464 *
***********************************************************************
------------------------------
Subject: Looking for simulator
From: klynn@ksuvax1.cis.ksu.edu (Kevin J. Lynn)
Organization: Kansas State University, Dept of Computing & Information Sciences
Date: 14 Nov 89 20:09:55 +0000
We are looking for a neural-net simulation system to run on an SCS
super-mini computer. The SCS uses the same instruction set as the CRAY
X-MP line and is running the CTSS operating system.
Any help appreciated. Please respond directly to us at the above address.
Thanks.
------------------------------
Subject: "Neuralsource" on disk?
From: gem.mps.ohio-state.edu!swrinde!cs.utexas.edu!asuvax!hrc!gtx!al@tut.cis.ohio-state.edu (Alan Filipski)
Organization: GTX Corporation, Phoenix
Date: 20 Nov 89 17:52:21 +0000
I just got a flyer from Van Nostrand Reinhold describing a book called
"Neuralsource-- The Bibliographic Guide to Artificial Neural Networks" by
Wasserman and Oetzel that they are selling for $64.95. It is a
bibliography of some 4000 references. They say that "this material was
originally available only on disk for computer retrieval"
Does anyone know if the disks are still available and from whom? I'd
prefer a computer file to a big book.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
( Alan Filipski, GTX Corp, 8836 N. 23rd Avenue, Phoenix, Arizona 85021, USA )
( {decvax,hplabs,uunet!amdahl,nsc}!sun!sunburn!gtx!al (602)870-1696 )
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
------------------------------
Subject: Test Data for NN
From: meo@kronos.cs.wayne.edu (Mike Otten)
Organization: Wayne State University, Detroit
Date: 20 Nov 89 19:35:57 +0000
Fellow Neural Netters,
I am working on implementing a neural net algorithm and am in search of
test data with which I can compare my results.
I have met with some success in training the net to classify sample
patterns from a 2-d XOR domain, and from a 7-d character recognition domain
with noise (LED digits). I am now looking for a standard
multi-dimensional, real-valued domain. I am looking for some
recognized/published domain and would prefer some synthetic dataset where I
can generate small sample sets for training and larger/more complete sets
for testing.
If some data bank can be made available to me or if you can offer any
pointers to what I am looking for, your help would be greatly appreciated.
Sincerely,
Mike Otten
Computer Science
Wayne State University
Detroit, MI 48201
313-577-2477
meo@cs.wayne.edu
------------------------------
Subject: SN2
From: hardbody@milton.acs.washington.edu (Hardbody)
Organization: Univ of Washington, Seattle
Date: 20 Nov 89 21:35:08 +0000
Has anybody out there used SN2 before? (The neural network simulation
package.) I need some help on it.
Please send e-mail.
------------------------------
Subject: Neural Network Simulator Available
From: max@shodha.dec.com (Max McClanahan)
Organization: Digital Equipment Corp. - Colorado Springs, CO.
Date: 27 Nov 89 20:10:28 +0000
Digital Equipment Corporation's Neural Design and Simulation
System (NeurDS) has recently been made available for use by
educational institutions. The NeurDS software is being licensed
on a no-fee basis to educational institutions who would like to
use it for neural network experimentation.
The NeurDS system is a general purpose tool for building, running
and analyzing Neural Network Models in an efficient manner.
NeurDS will compile and run virtually any Neural Network Model
using a consistent user interface that may be either window or
"batch" oriented. The NeurDS system will run on any Digital
platform including Vax/VMS, Vax/Ultrix, and DECsystem/Ultrix. A
graphics terminal is not required to support the window interface.
The system includes a users manual as well as examples of several
standard neural network problems.
Copies of the NeurDS system may be requested by writing to:
mcclanahan%cookie.dec.com@decwrl.dec.com
Digital Equipment Corporation
1175 Chapel Hills Drive
Colorado Springs CO 80920-3952
------------------------------
Subject: neural net software wanted
From: westby@agnes.acc.stolaf.edu
Organization: St. Olaf College, Northfield, MN
Date: 27 Nov 89 21:53:51 +0000
A friend of mine is looking for information on neural net software,
especially programs written for an IBM-PC. If you know of any such
programs, or have references to somebody who knows a lot about these
programs, please send e-mail to me at
westby@thor.acc.stolaf.edu
-or-
westby@agnes.acc.stolaf.edu
Thank you very much.
Ina Westby
westby@thor.acc.stolaf.edu
westby@agnes.acc.stolaf.edu
------------------------------
Subject: GMU bps update
From: enorris@gmuvax2.gmu.edu (Gene Norris)
Organization: George Mason University, Fairfax Va.
Date: 28 Nov 89 00:14:06 +0000
An updated version, 1.01, of the Goerge Mason University back prop
simulator, bps is now available via anonymous ftp from gmuvax2.gmu.edu.
This is a Unix system. The file bps.tar is a tar-formatted compressed
version of the GMU backprop simulator, tutorial(s) and sample files. In
order to use it, you must tar it to de-archive it and then uncompress the
resulting 8 files. When you get that far, please read the brief file
readme.now before proceeding further.
The current version, 1.01, of bps was released on Tuesday, Nov. 14 and
contains a fix to the transfer function derivative that was incorrect if
you used a symmetric transfer function, i.e. one with a range of -M/2 to
+M/2. AT present, the bug fix is only in the unix version -- we've had
local network problems as well as no time. Please bear with us on this.
When you run the updated version, the version number appears on the initial
screen.
A SOURCE CODE license is now available. Please contact the undersigned for
details .
Eugene Norris
CS Dept George Mason University Fairfax, VA 22032 (703)323-2713
enorris@gmuvax2.gmu.edu
------------------------------
Subject: new book announcement
From: Steve Hampson <hampson@ICS.UCI.EDU>
Date: Tue, 28 Nov 89 13:36:25 -0800
New Book Announcement!
Connectionistic Problem Solving
Steven Hampson, UC Irvine
Birkhauser: Boston
ISBN 0-8176-3450-9
276 pages, $39.00
is now available from:
Birkhauser Boston, Inc. Order Dept.
c/o Springer Verlag
44 Hartz Way
Secaucus, NJ 07094
1-800-777-4643
Here is the back-cover blurb.
This book develops a connectionistic model of adaptive problem solving. In
the process, various computational issues are identified that are
susceptible to formal analysis and that can be related directly to the
biological structures and processes involved in biological problem solving.
Abstract problem solving is a well-defined domain for the formal
investigation of intelligent behavior, and maze learning, used here as a
concrete example of problem solving, is a traditional tool for
psychologists and neuroscientists studying animal behavior. Consequently, a
substantial body of physiological and behavioral data is available to
motivate the construction of computational models. Conversely, functional
issues arising in model building help organize these often disparate
biological results. At the most general level, the book addresses issues
concerning the efficient learning and representation of categories, and the
assembly of action sequences to achieve specific goals. Various
approaches, often with complementary characteristics, are considered. An
extensive bibliography, ranging from the work of Tolman to recent
connectionist theory, provides access to relevant work in the
neural-modelling and animal learning-theory literature.
------------------------------
Subject: Neural Networks on Transputers
From: hfry@alize.imag.fr
Organization: IMAG-TIM3/INPG, Grenoble, France
Date: 13 Oct 89 14:44:51 +0000
I am trying to implement NN algorithms on a MegaNode (128 Transputers)
and I am looking for articles, abstracts or any references on attempts made
on parallelisation of NN.
Thank you
Herve Frydlender
Laboratoire TIM3 APCF
INPG
46 avenue Felix Viallet
38031 GRENOBLE Cedex
FRANCE
e-mail : hfry@alize.imag.fr
------------------------------
Subject: Connectionist models of music
From: derek@prodigal.psych.rochester.edu (Derek Gross)
Organization: University of Rochester Cognitive Science
Date: 07 Oct 89 22:32:01 +0000
Does anyone know of any connectionist models of musical structures, perception
or composition? If so, please send me e-mail.
Thanks,
Derek Gross
------------------------------
Subject: musical applications of neural nets
From: viseli@uceng.UC.EDU (victor l iseli)
Organization: Univ. of Cincinnati, College of Engg.
Date: 13 Oct 89 18:42:02 +0000
Over the past few months, there have been various requests (mine included.)
for information on music-related applications of neural nets. Since I have
not seen any mention of it yet, I felt that I should bring to light a
newly-found source which might contain some valuable articles:
Computer Music Journal, vol. 13, no. 3.
"Neural Nets and Connectionism 1"
This journal is published by the MIT Press, and subscription in-
formation is available at
The MIT Press
55 Hayward Street
Cambridge, MA 02142.
A brief run-down of the rates ... "$25 for students and retirees, $28 for
individuals, and $63 for institutions. Subscribers outside the United
States add $9 for surface postage, $17 for airmail. Single copies of
current issues: $8." Issues appear quarterly.
This particular issue, the fall issue, contains articles covering
-- "Preface to the Special Issue on Parallel Distributed Processing
and Neural Networks"
-- "Machine Tongues XII: Neural Networks"
-- "A Neural Net Model for Pitch Perception"
-- "Connectionist Models for Tonal Analysis"
-- "The Quantization of Musical Time: A Connectionist Approach"
-- "Using Connectionist Models to Explore Complex Musical Patterns"
-- "Fingering for String Instruments with the Optimum Path Paradigm"
-- others including more special interest and updates on new products.
This is the first issue of a two-issue Neural Network series being run by
this journal. The next issue of Computer Music Journal is expected in the
Winter (Dec./Jan.), and will expand on some of the subjects presented in
the first featured neural network issue. If you are interested in neural
nets in music, I suggest checking this publication out.
- ------------------------------------------------------------------
|| || Victor Iseli (viseli@uceng.uc.edu)
|| University of Cincinnati
\\ // || || Dept. Electrical Engr.
\\ // || || 811K Rhodes Hall
\\// || || Cincinnati, OH 45219
- ------------------------------------------------------------------
------------------------------
Subject: Re: MUSIC and Neural Nets
From: schiebel@a.cs.wvu.wvnet.edu (Darrell Schiebel)
Date: 17 Nov 89 22:33:50 +0000
I'd suggest getting the paper by Teuvo Kohonen entitled "A Self-Learning
Musical Grammar, or `Associative Memory of the Second Kind'"; it was
published in the International Joint Conference on Neural Networks June
18-22, 1989. The music generated was quite beautiful.
BTW: Teuvo Kohonen, Helsinki University of Technology, Laboratory of
Computer and Information Science, Rakentajanaukio 2 C, SF-02150
Espoo, Finland.
Darrell Schiebel
(schiebel@a.cs.wvu.wvnet.edu)
[[ Editor's Note: I was actually quite disappointed both with the paper and
the music. It sounded to me even less interesting than Mozart's Musical
Dice results. Admittedly, being classically trained at Conservatory, I
have a somewhat biased opinion. Perhaps to more generous ears, the music
was adequate. -PM ]]
------------------------------
Subject: connectionist models of music
From: todd@psych.Stanford.EDU (Peter Todd)
Organization: Stanford University
Date: 29 Nov 89 03:28:18 +0000
In answer to recent queries on the net about connectionist models of music,
I wanted to point out that the current and next issues of the Computer
Music Journal (MIT Press) are specifically devoted to this topic. In the
current issue, 13(3), out now, there is a general tutorial on musical
applications of networks, plus articles on network models of pitch
perception, tonal analysis, quantization of time, complex musical patterns,
and instrument fingering. In the next issue, 13(4), due out at the end of
the year, there will be articles on my work using sequential networks for
composition, modelling tonal expectancy (with Jamshed Bharucha, who has
also published much work in the area of network modelling of human musical
behavior), and another article on representations for pitch perception.
Both issues were edited by D. Gareth Loy, of UC San Diego, and myself; the
journal is available in some bookstores. Hope this helps--
peter todd
stanford university
psychology department
todd@psych.stanford.edu
------------------------------
Subject: Re: MUSIC and Neural Nets
From: eru!luth!sunic!mcsun!ukc!reading!cf-cm!andrew@bloom-beacon.mit.edu (Andrew Jones)
Organization: University of Wales College of Cardiff, Cardiff, WALES, UK.
Date: 27 Nov 89 18:12:42 +0000
I would like to thank all those who have e-mailed responses to my original
request concerning musical applications of Neural Networks. Each message
was very welcome. I have replied to each message I have received, so
hopefully the mailers are already aware of my gratitude.
Some of the mailers asked if I could summarize responses to the net. A
number of people pointed me towards the work of Teuvo Kohonen. He has
published the following paper:
A Self-Learning Musical Grammar, or 'Associative Memory of the Second Kind'
International Joint Conference on Neural Networks, 1989, Vol 1, p. I-1 to
I-5.
Apparently the music generated has something of the flavour of J.S. Bach.
I understand that Teuvo Kohonen will present a "Neural Concert" on Monday
Jan. 15th, 1990 at the 1990 IJCNN. I wish I could be there!
Apart from this, the Computer Music Journal has devoted Vol. 13 nos. 3 & 4
to this topic.
Rick Romero (rr2p+@edu.cmu.andrew) stated that he is considering getting
involved in this area; Andrew McCallum (andrewm@edu.dartmouth.eleazer)
informed me that Jamshed Bharucha, Prof. of Psychology at Dartmouth
College, has been doing research into NN's and the perception of music for
quite a while. He can be reached at bharucha@dartmouth.edu.
Finally, I have not yet received a posting by Eliot Handelman, except that
Darrell Schiebel included at least part of it in his posting (things are V
E R Y S L O W getting through to us at the moment). Eliot Handelman is
saying, I think, that non-musicians should avoid this subject area. One of
the people who e-mailed me stated that he disagreed with this opinion. I
suppose it depends what you mean by "musician".
Anyway, thanks again to all those who expressed an interest!
ANDREW
======
------------------------------
End of Neurons Digest
*********************
