Page 129 - Biomimetics : Biologically Inspired Technologies

Bar-Cohen : Biomimetics: Biologically Inspired Technologies DK3163_c003 Final Proof page 115 21.9.2005 11:41pm

Mechanization of Cognition 115

Only two of the answer lexicon symbols shown in Figure 3.A.6, namely e and l L are receiving
links from all four assumed facts. However, note that the links impinging on the neurons of symbol
e are stronger than those impinging on symbol l L . Now this discussion of the biological imple-
mentation of confabulation will pause momentarily for a discussion of synapses.
Despite over a century of study, synapse function is still only poorly understood. What is now
clear is that synapses have dynamic behavior, both in terms of their responses to incoming action
potentials, and in terms of modifications to their transmission efficacy (over a wide range of time
scales). For example, some synapses seem to have transmission efficacy which ‘‘droops’’ or
‘‘fades’’ on successive action potentials in a rapid sequence (such are sometimes termed depressing
synapses — which has nothing to do with the clinical condition of depression). Other synapses
(termed facilitating) increase their efficacies over such a sequence; and yet, others exhibit no
change. However, it has been learned that even these categorizations are too simplistic and do not
convey a true picture of what is going on. That clear picture awaits a day when the actual
modulations used for information transmission, and the ‘‘zoo’’ of functionally distinct neurons
and synapses, are better understood. Perhaps this theory can speed the advent of that day by
providing a comprehensive vision of overall cortical function, which can serve as a framework
for formulating scientific questions.
Even though little is known about synapses, it is clear that many synapses are weak (unstrength-
ened), quite likely unreliable, and marginally capable of signaling (this theory claims that roughly
99% of synapses must be in this category, see Section 3.A.7). This is why it takes a pool of
highly excited or active neurons representing a symbol (such neurons possess the ultimate in
neural signaling power) to excite transponder neurons (each of which receives many inputs from
the pool). No lesser neural collection is capable of doing this through unstrengthened synapses
(which is why cortical synfire chains have only two stages). However, it is also known that
some synapses (this theory claims that these represent fewer than 1% of the total of cortical
excitatory synapses, see Section 3.A.7) are much stronger. These stronger synapses (which
the theory claims are the seat of storage of all cortical knowledge) are physically larger
than unstrengthened synapses and are often chained together into multiple-synapse groups that
operate together (see Figure 3.A.7). One estimate (Henry Markram, personal communication) is
that such a strengthened synapse group can be perhaps 60 times stronger than the common
unstrengthened synapse (in terms of the total depolarizing effect of the multi-synapse on the
target cell at which they squirt glutamate neurotransmitter). These strong synapses are probably
also much more reliable. Figure 3.A.7 illustrates these two hypothesized types of cortical excitatory
synapses.
The theory hypothesizes that synapses which implement knowledge links (as in Figure 3.A.5)
are always strengthened greatly in comparison with unstrengthened synapses. When the knowledge
link requires that a transponder–neuron-to-target–symbol–neuron synapse code the graded prob-
ability p(cjl) (as opposed to just a binary ‘‘unstrengthened’’ or ‘‘strong’’), the dynamic range of
such a strengthened synapse is probably no more than a factor of, say, 6. In other words, if the
weakest strengthened synapse has an ‘‘efficacy’’ 10 times that of an unstrengthened synapse,
the strongest possible synapse will have an efficacy of 60. Thus, we must code the smallest
meaningful p(cjl) value as 10 and the strongest as 60.
In our computer confabulation experiments (e.g., those reported in Hecht-Nielsen, 2005
and many others), the smallest meaningful p(cjl) value (define this to be a new constant
p 0 ) turns out to be about p 0 ¼ 0.0001 and the largest p(cjl) value seen is almost 1.0. As it
turns out, the smaller p(cjl) values need the most representational precision; whereas little
error is introduced if the larger p(cjl)’s are more coarsely represented. Clearly, this is a situation
that seems ripe for using logarithms! The theory indeed proposes that nonbinary strengthened
synapses in human cortex have their p(cjl) probabilities coded using a logarithmic scale (i.e., y ¼
log b (cx) ¼ a þ log b (x), where a ¼ log b (c)). This not only solves the limited synaptic dynamic