I was curious to what the feedback to this model of
consciousness would be. The model is definitely based on a desire to
link brain and consciousness but consider the ideas present
regardess of whether you think consciousness emerges from
brain processes or not.
I think he raises some undiscussed issues like
expectation, intentionality, unconscious procedural memories that
never resonate.
There is a lot of jargon in it but I would appreciate any feedback
about the model.
Thanks,
Erin
Grossberg's ART model:
The mind solves a very hard problem that many current approaches to technology
have not solved. It is a self-organizing system that is capable of rapid yet
stable autonomous learning of huge amounts of data in a nonstationary
environment. The processes wherby our minds continue to learn about a changing
world in a stable fashion throughout life are proposed to lead to conscious
experiences. These processes include the learning of top-down expectations,
the matching of these expectations against bottom-up data, the focusing of
attention upon the expected clusters of information, and the development of
resonant states between bottom-up and top-down processes as they reach an
attentive consensus between what is expected and what is there in the outside
world. Grossberg calls the problem whereby the mind learns quickly and stably
without catastrophically forgetting its past knowledge the
stability-plasticity dilemma. Grossberg’s solution to this dilemma is ART
(adaptive resonance theory). ART predicts that all conscious states are
resonant states.One also needs to respect the fact that our minds can adapt on
their own to changing environmental conditions without being told that these
conditions have changed (intelligent being can autonomously adapt to a
changing world). Knowing how to do this is presently an art form. There are
no known algorithms with which to point the way.
The process whereby top-down expectation selectively amplifies some features
while suppressing others help to “focus attention” upon info that matches our
momentary expectations. This focusing helps filter signals so we aren’t
overwhelmed and prevents destabilizing of previously learned memories. This
reciprocal feedback of top-down expectations and bottom-up signals takes
awhile to equilibrate but when it does it locks the short-term memory activity
patterns of the interacting levels into a resonant state. Only resonant
states can drive new learning. ART suggests that stability implies
intentionality. Intentionality has two senses; anticipation of events and
attention and consciousness (focus on data worthy of learning).
Implicit in the concept of intentionality is the idea that we can get ready to
experience an expected event so that when it finally occurs we can react to it
more quickly and vigorously and until it occurs ignore less desired events.
The top-down signals represent the mind’s learned expectations of what the
bottom-up signal patterns should be based on upon past experience. The
matching process is designed to reinforce and amplify those combination of
features in the bottom-up pattern that are consistent and suppress those
features that are inconsistent.
If the mismatch b/w bottom-up and top-down is too great for resonance to
develop the category is reset and a memory search for a better category is
iniated. Whether a resonance occurs depends upon the level of mismatch or
novelty that the network is prepared to tolerate. Novelty is measure by how
well a given exemplar matches the prototype that its pres evokes. The
criterion of an acceptable match is defined by an internally controlled
parameter called vigilance. Low vigilance leads to broad generalization and
abstract prototype. High vigilance leads to narrow generalization and to
prototypes that represent fewer input exemplars or even single exemplars.
Thus the contents of a conscious percept can be modified by environmentally
sensitive vigilance control.
It is also suggested that sensory and cognitive processing in the
What processing stream of the brain obey top down matching and learning laws
that are often complementary to those used for spatial and motor processing in
the brain's Where processing stream. This enables our sensory and cognitive
representation to maintain their stability as we learn more about the world
whi while allowing spatial and motor representations to forget learned maps
and gains that are no longer appropriate as our bodies develop and grow from
infanthood to adulthood. Procedural memories are proposed to be unconscious
because the inhibitory matching process that supports these spatial and motor
processes cannot lead to resonance.
Erin
MOQ.ORG - http://www.moq.org
Mail Archive - http://alt.venus.co.uk/hypermail/moq_discuss/
MD Queries - horse@darkstar.uk.net
To unsubscribe from moq_discuss follow the instructions at:
http://www.moq.org/md/subscribe.html
This archive was generated by hypermail 2b30 : Sat Aug 17 2002 - 16:02:20 BST