ABSTRACT

S. Dehaene and J. P. Changeux’s model was intended to represent more general cognitive and inferential capabilities than are manifested by the Wisconsin card sorting test, which is a simple test for mental flexibility in the face of changing context. As brain imaging results supplement animal, lesion, and behavioral studies to provide more pieces of the cognitive neuroscience puzzle, the goal of a self-consistent computational model of the entire brain seems inviting and possibly attainable. Researchers who develop models of complex brain function vary widely in their relative emphasis on neuroscientific versus behavioral or cognitive details. The use of spikes for synchronization of responses to inputs to model perceptual grouping has been a feature of many other visual cortex models. Bayesian models have been applied to a wide range of cognitive processes including reasoning, reinforcement learning, and many others.