Re: face-selective region

"feedbackdroids" <feedbackdroids@xxxxxxxxx> wrote:
Curt Welch wrote:

Well, yes, we should expect his simplified model to be too simplified
for a full and complete description of the brain (he even says so in
the book), but the evidence you mention above fits exactly with his
predictions so I don't really know why you think it's an indication of
addition complexity.


My point was there are probably multiple areas with face cells, not
just one. The more people look at things, the more complexity they
find. Hubel+Wiesel found 3 or 4 visual areas with retinotopic maps,
then later Kaas et al found 20 or whatever, and others just kept adding
more with further study. Now 30+.

Yeah, as expected. If Jeff's type of idea is at all correct, then you
would expect to find nearly infinite complexity if you attempt to find are
borders because they would only be loosely defined.

A few people such as Desimone and Ungerleiter are beginning to get a
feel for how to control for activity in the myriad feedback pathways
between cortical regions, and as this proceeds, things should really
start to get interesting.

Yeah, the one thing that Jeff's book shows is that we are very near to
understanding exactly what the cortex is doing (near in terms of decades at
least).

I imagine the summary is that some cortical regions have
very-specific functions, while in others you find more heterogeneity,
with cell responsiveness covering a wider range of stimuli. This
should be a good approach where you need to do 2 things: (a) compute
specific functions or operations, and (b) combine together the
results of many such operations.

All Jeff's model seems to say (as I read it) is that the entire cortex
is a invariant pattern recognizer. This means that every micro column
in the cortex (both motor and sensory cortex) will become active and
stay active whenever the external condition it's responsible for
recognizing is present. His model makes little to know predictions
about how those pattern recognizers will group together to form similar
cortical regions - other than that similar patterns are more likely to
be located close together in the cortex. All this means is that 1)
there should be face recognizing columns in the cortex and 2) you would
expect to find face recognizing columns close to other face columns.


My point was there are probably some areas which perform very specific
computations while other areas perform what used to be called
associational [ie, multimodal integration, etc] activities.

Yeah, well, I like Jeff's type of approach which basically says that all
areas are doing multimodal integration. Some just happen to be working
withing the same sensory domain because they are integrating data from
different sensors from the same same modality.

I actually agree with most of Jeff's ideas, except he seems to have
missed what I see as the most important half of the problem. His ideas
seem to align with my old INET concept - an unsupervised learning
system for identifying the current state of the environment. But he
makes no mention of what was the ONET side of my approach - which is
the part which takes the environment state as identified by the INET,
and maps it to actual behavior, using a reinforcement learning
algorithm. As far as I can tell, his theory fails to answer the single
most important and most fundamental AI problem - how does the brain
decide what behavior to produce.


Not sure why you say this referencing your ONET, but Hawkins' model has
upwards directed hierarchies on the sensory side and downwards directed
hierarchies on the motor side. IOW, the model is quite symmetrical on
both sides. The same sort of processes that lead to successive
abstraction on the sensory side produce the inverse [ie, more and more
specifics leading to muscle behavior] on the motor side. Essentially
inverse abstraction.

True. That's the exact same idea that drove me to the INET ONET idea. And
you are right, as Jeff talks about it, he does have a symmetric system like
that. However, he's wrong. It's because as far as I can tell, he never
explains how the motor cortex manages to perform the inverse function of
the sensory cortex if it's the same hardware and the same algorithm.
Hardware doesn't just "invert" like that magically. And even if there was
some basic DNA switch that caused it to be wired differently to create the
inverse function, he doesn't even touch on the problem of how it manages to
make the inverse decisions. For example, he gives an example of the motor
cortex creating an invariant representation of the Gettysburg Address, and
then when translating that into action, it can either send the signals to
the fingers to type the message, or send it to the signals to the mouth to
speak the address. But what criteria would the brain working under his
model use for making that decision? How does the brain make the decision
to type the Gettysburg Address instead of speaking it? This is what you
need a reinforcement learning systems to answer and this is the huge and
most important question about how the brain works that he seems to just
ignore (or wash over by pretending his memory prediction framework would
just magically produce intelligent behavior without having to answer how
these decisions were made).

This is where I believe he would appreciate my theory about the motor
cortex. My theory is that the motor cortex isn't in fact the inverse of
the sensory cortex - it's doing the exact same thing as the sensory cortex.
However, instead of being sent sensory signals, it's sent the motor outputs
from the lower brain. So the purpose of the motor cortex is not to
generate motor behavior, but to sense it. So the motor cortex is in fact
just yet another sensory cortex, sensing the behavior of the lower brain.

In my theory, this would then imply that the entire sensory and motor
cortex, was then responsible for modifying the behavior of the lower brain.
But for this to be true, the brain would have to have outputs from the
entire cortex, leading to systems that allowed it to modify motor signals.
But surprise surprise, Jeff says that this is exactly what is happening in
the cortex - that the entire cortex sends outputs to the lower brain to
modify outputs. He talks for example about how the primary vision system
is known to be controlling eye movements.

So where I was 90% certain my theory about the motor cortex just being more
sensory cortex before reading Jeff's book, I'm not 98% certain that is
exactly what is happening.

The point of the entire cortex, the sensory, and motor cortex, is to take
raw sensory data, and use that to define the current invariant state of the
environment. So if there's a face in the environment, there will be
invariant activity somewhere in the cortex to indicate that there's a face
in the environment.

From there, you have to add a reinforcement learning system to map the
current state (i.e., the stimulus signals), to behavior. So somewhere in
the link from the state defined by the activity in the cortex to behavior,
the brain must be implementing a reinforcement learning system that
strengthens behaviors based on rewards and weakens behaviors based on
punishments.

Regards which behavior is to be ultimately produced, well, H's model is
as yet incomplete. It's basically a framework for a model, with many
details needing to be filled in. Likewise, I thought it was a little
too weak on temporal aspects, production of multiple sequential
operations, etc. I should probably review these aspects of H.

His idea of how it generates output sequences is that it performs the
inverse of the sensory system. So what needs to happen is that if you
wanted to create a sequence like what was required for the Gettysburg
Address, the high level system needs to first activate the cells which
represent the concept of the Gettysburg Address. Those would activate and
stay active for the entire processes. Those in turn would trigger the next
lower concepts in the sequence, like "Four". That would stay active for as
long as you were saying or writing "Four", and would in turn generate the
next lower level sequence of signals needed to produce the "Four" behavior.
The "Score" behavior would be triggered by the combination of context
signals "Gettysburg Address" and "Four", and maybe even the the lower level
signals for the trailing behavior of producing the "Four" sound. So, the
only temporal signalling needed in the hardware is very short term - just
to create the correct delay from the end of "Four" to the beginning of
"Score". So the hardware which activates the "Score" internal state only
needs a short term temporal relationship to the context that exists just
before we would say "Score" when trying to generate the entire sequence
known to us as the Gettysburg Address.

Each step of a long sequence is generated by the hardware for each step
recognizing that it was now time to become active based on all the context
signals active in the cortex.

Where Jeff seems to see this as the motor cortex doing some inverse of what
the sensory cortex is doing, it doesn't in fact have to do that at all. It
can do the same thing just by acting as another sensory cortex watching
what outputs the lower brain is sending to the muscles. By doing that, the
motor cortex "knows" in effect, the current context of the motor behavior.
The motor cortex is in effect recognizing that we are currently reciting
the Gettysburg address. And using that fact, each next behavior in the
sequence can be triggered, based on how far we have currently advanced in
the processes. So the motor cortex is driving behavior, but it's driving
it by telling the rest of the brain where we are in the processes at each
moment.

Rememer H had the hippocampus at the "top" of his hierarchy, but I
think this was just a wild shot in the dark, and wasn't too compelling.

Yeah, I wasn't either.

I suspect the correct answer is that the hippocampus is the controller in
the reinforcement learning system. It's the hardware which implements
reinforcement learning. And the reason Jeff couldn't figure out what it's
purpose was is because Jeff hasn't yet grasped that there must be a
reinforcement learning system at work to explain how the brain actually
shapes behavior, instead of just shaping knowledge - which is what his half
of the equation answers.

However, with such a massively-interconnected system, you might not
need a top-level executive at all. Rather, recurrent activity in the
feedback loops keeps driving one thing to the next.

Right, as long as you have some type of temporal system that can active
signal X some short but fixed period of time after signals Y and Z happens,
then you have the foundation for a sequence generator. Whether that delay
is created with a recurrent feedback loop or other system is not important.

The output of each
process is the trigger to the next one in sequence,

Exactly.

but at the same
time, the overall activity is continually being refined by the same
recurrent activity. The loop activity is continually stimulating output
from the associative memory, which enters back into the loop activity.

But to create our type of structured behavior, you also need signals to
define long term context - such as the signal that activates, and stays
active, the entire time you are reciting the Gettysburg address. This is
needed to allow you to say Four Score in that context, and Four Five, when
in the context of reciting your numbers. Without a hierarchy of context
signals to guide the sequence generation, you would be forced to always
follow the word "four" with the same behavior. So you could learn to say 4
5, or four score, but not both.

And that is exactly what the neocortex is doing for us. It creates a
hierarchy of context signals not otherwise present in the raw sensory data.
Where as there is no single signal in the raw sensory data to let the
behavior generation system know there is a cat in front of us, the cortex
learns to spot cats and gives us that "cat" signal so we can generate
different output sequences when there is a cat in front of us than when
there is a dog in front of us.

One criticism of behaviorism is that behaviorism doesn't tell us how the
system learns to "see a light" as a stimulus signal. Jeff's theory about
how the cortex operates as an invariant signal generator answers that for
us. You just tack on a reinforcement learning system to create the correct
mapping from invariant signals to output behaviors, and feedback the final
output to the same invariant signal generator to allow the motor cortex to
create invariant signals which describe what we are doing, and there you
have the correct and full framework to answer what the brain is doing.

I think the entire system works just like the example given several
times in the past of the tennis player hitting the ball over the net.
Things just don't start and stop, and decisions made, at "specific"
points in time [like Libet tried to measure in his experiments].
Rather, one thing leads to and blends into the next, and each change in
input and output that occurs in the meanwhile refines the onging
activity. A guy just doesn't hit the ball when it gets to the racket.
Rather, he continually follows the ball visually, moves his body into
position predictively, prepares the racket for where he will move it
to, on and on. There is no one process and one specific time point,
there are many things going on constantly which are interacting with
and modifying each other. Same for tennis, same for language, same for
other sequential activities. This goes back also to Clark's book we
mentioned recently that the brain evolved to control actions.

yeah, exactly. The behavior generated by the brain is just a constant flow
of micro behaviors all triggered by the current context as understood and
decoded by all the signals in the cortex. Ever pulse generated, is the
next micro behavior. That's the level of behavior the brain is working
with. Pulses. Each pulse is the brains next behavior. So, in order to
hit a ball with the racket, the brain must generate millions of pulses
(micro behaviors) all based on the current context of the environment, as
defined by the sum total of all the activity in the neocortex.

And I've now moved on to a solving both problems in each node instead
of having to use two networks to do it. My pulse sorting net attempts
to solve both sides of the problem at once.

His predictive memory approach explains in fairly detailed speculation,
how the brain "knows" what is going on in the environment and how it
"knows" how the world works, but totally fails to say anything about
how the brain uses that knowledge to make behavior decisions.


As mentioned above, H does have a downward directed motor hierarchy
which is complement to the sensory side, along with associational
connections on top of the hierarchy.

Right, he does. Right idea, but wrong implementation. The motor cortex is
not an inverse of the sensory cortex. Motor outputs flow up exactly the
same as sensory data flows up in the cortex.

It sounds like your new nets have abandoned a hierarchical scheme
altogether. No?

Hell no I've not abandoned it. The hierarchy is the foundation of our
complexity. Without a hierarchy at work we would not be able to understand
that we were listening to the gettysburg address instead of just listing to
a sequence of words. Without a hierarchy at work, we couldn't produce the
word "four" in two different contexts. We would have to learn the word
once to say it in the sequence "four five six", and learn the behavior of
speaking the word a second time from scratch for the sequence "Four score
and ..".

My network is the hierarchy. Each level of network adds another layer to
the hierarchy. The function of a node in the middle of the network is in
effect "used" by all the upstream nodes that have the power to route data
to it, and in turn, all the downstream nodes that can receive a pulse from
a node in the middle of the network are being "used" by the node. It's
very much a huge hierarchy. And when you add global feedback to create the
motor cortex half of the system, you in effect create a hierarchy of
infinite depth (not to mention that the same thing happens just because of
the feedback though the environment that happens (we listen to ourselves
talk and that creates infinite recursion limited only by time itself).

So, for example, his
theory might explain how a mouse or a human, would recognize part of a
hallway and know before turning the corner, that the cheese is probably
right around the corner, but no where does he seem to answer why the
brain would try to get the cheese, instead of trying to run away from
it.


Oh, goals and emotions. As I recall, H specifically say emotions
shouuld best be left out of this model, in favor of having a completely
rational system [eg, don't want a air traffic-controller AI freaking
out - [aside, ever see Pushing Tin?].

Emotions are the correct answer. That's exactly what a reinforcement system
creates. But without them, you have no purpose, no drive to do anything,
nothing to make you care about eating the cheese instead of smashing it
with a hammer. Without emotions, you have no purpose, and without that,
you have no intelligence. Understanding the world doesn't help if you
don't have something to do with the understanding. You can't build a
rational AI system. That's an oxymoron. If it's not emotional, it ain't
intelligent.

If you want to build a mechanical reasoning engine, then you can leave out
the emotion, but you also leave out the intelligence and common sense when
you do that. There's a place in the world for such systems, but it's not
real AI.

The actual brain has several sub-cortical systems that work in
"parallel" to act on the cortex, to do with goals and emotions. Edelman
called them value systems, I believe. I'm not sure what are H's ideas
on this.

Yeah, emotions, and values are all the same. The heart of all reinforcment
learning system is value calculation. It must calcuate a value fuction
which is used to select beahviors. It's what allows the system to pick one
beahvior over another. It's what causes the mouse to eat the cheese
instead of ignoring it.

Or,
even if it for some reason "wanted" the cheese, now the brain would
manage to learn the correct sequence of outputs to get the cheese into
it's stomach. There's where you need a reinforcement learning system.


You have your system here embodied and interacting with the
environment. H doesn't seem to really consider this, I don't think. Not
sure.

I'm sure he considers it. But mostly, his ideas are about how the system
understands the sensory data and he just hasn't talked much about how it
learns to prefer one behavior over another which like I say, is the real
problem of AI. It must "value" one behavior over another using some
mechanical system. He explains how important memory and prediction is to
understand how the world works, both for sensing and behavior, but he just
never seems to talk about how he expects his system to value it's
behaviors. So it understands that its hand will fall off if you whack it
with a butchers knife, but what makes the brain decide to not hack off its
own hand? He just doesn't seem to answer that most important of questions
with his theory.

I've not finished the book so maybe he talks about this later but I
seem to already be past his technical description of his theory.

So, it looks like he's on what I see as the right track, but has only
so far, addressed half of the problem.


Well, I think we all 3 agree on this, H being the third party.

yeah.

And of course, what I like about his theory is that he, like I, believe
there is some fundamental learning algorithm at work for the entire
cortex, and that specialized sections of the cortex form more because
of what data is sent to it, then the way the DNA built it.


Well, I think these last 2 are one and the same.

Sure, because it's the DNA that determines what data is sent to which part
of the cortex. The DNA has the upper hand in the end. But what this tells
us that that the algorithm we have to build is one and the same for the
entire cortex, and the topology is what we have to get right to create all
the different modules that are needed to equal human processing skills.

There's also no doubt that the DNA has also made optimizations to different
parts of the cortex to maximize it's performance for each type of data
processing that happens in the section.

The DNA just decides what
data gets sent to which part of the cortex, (i.e., where each sensory
signal is connected to the cortex and maybe how different sections of
the cortex are bulk-wired to other sections of the cortex) then the
function formed in each section is a result of the characteristics of
the data that shows up there. So the DNA defines the network topology,
and the generic learning hardware which is the function of the cortex,
auto configures itself to understand the type of data it is dealing
with.


Well, of course, this last bit is one of the big mysteries. ;-)

It can't be quite as loosey-goosey as you say here, else you wouldn't
have the same areas doing the same things in every different member of
the species.

Sure you would. The mammals all share common ancestors so they also share
common design traits. If you hook the optic nerve up to the back of the
brain in a mouse and in a human, you would expect find the visual centers
in the back of the brain in both a mouse and a human.

If mice and humans tend to look at the same world, with a similar type of
eye, then you would expect the data flowing to the visual cortex to be very
similar in nature as well and that would lead to very similar hierarchies
of understanding developing (both human words and mice words have lots of
constant objects generating consistent visual patterns to be recognized).

--
Curt Welch http://CurtWelch.Com/
curt@xxxxxxxx http://NewsReader.Com/
.

Relevant Pages

  • Re: Neural netss (was Re: death of the mind.)
    ... > In the cerebellum the output signals for motor control are comparator ... The way the entire percept in the cortex ... > switches in binocular rivalry would be expected from the feedback from ... >> The reason I keep talking about brain physiology as I understand it is ...
    (sci.cognitive)
  • Re: face-selective region
    ... current state (i.e., the stimulus signals), to behavior. ... Inverse hierarchy. ... signals active in the cortex. ... what outputs the lower brain is sending to the muscles. ...
    (comp.ai.philosophy)
  • Re: We Robot
    ... Consciousness without a cerebral cortex: ... that consciousness is actually in the gray matter of the brain as it were.. ...
    (talk.origins)
  • Re: Strong AI Thesis (No Chinese room, I promise)
    ... The result showed the part of the cortex ... This is why whey they do brain ... The lower brain is not generic learning hardware. ... The higher level behaviors are more variable than the lower ...
    (comp.ai.philosophy)
  • Re: face-selective region
    ... cortex will become active and stay active ... using a reinforcement learning algorithm. ... AI problem - how does the brain decide what behavior to produce. ... Rememer H had the hippocampus at the "top" of his hierarchy, ...
    (comp.ai.philosophy)