A Plastic Temporal Brain Code for Conscious State Generation
Figure 1
Genesis of resonance states in a dedicated circuit with five principal resonant neurons acting as “coincidence detectors.” Figure 1(a) illustrates a dedicated resonant circuit with five principal resonant neurons acting as coincidence detectors. Each apex of a given polygon corresponds to a principal resonant neuron which can receive input or emit output signals from and to processors anywhere in the brain, along the long-distance tracks of resonant circuitry that has been primed in the course of brain development to generate the temporal activity patterns for conscious state generation. Unidirectional priming only is shown here as one possible example, for illustration. Each edge of a polygon represents a delay path which transmits signals from a given apex to the next, with a characteristic delay that would correspond to some multiple of the elementary “bin” unit. All principal resonant neurons would have been primed throughout lifespan brain development to preferentially process input which carries statistically “strong” signals. When activated, principal resonant neurons send signals along all delay paths originating from them, and all those receiving a signal coinciding with the next input signal remain activated. The connections between principal resonant neurons of such a model would thus be potentiated as in the classic Hebbian model. Figure 1(b) shows some of the many possible excitation patterns within a dedicated resonance circuit with only five principal neurons. Such circuits would form interconnected neural networks that extend across large distances across the brain and have intrinsic, essentially arbitrary though not random, topologies in terms of “which cell fires first.” Such intrinsic topology is unrelated to functionally specific spatial cortical maps. As in the world some events are more likely than others, the same holds for brain events. Whether a given temporal resonance pattern will or will not generate a conscious state is determined by statistical likelihood computations in the brain. How such computations may work is simulated in ART (e.g., [82]), or the TEMPOTRON model by Gutig & Sompolinski [118].
