The first of my two major research areas is the representation of spatial information in the rodent brain. I develop computational models of the hippocampus, an area thought to be involved in navigation and learning. There is evidence that the hippocampal "cognitive map" combines visual landmark information with another, internal sense of position maintained by path integration. I also model the rodent head direction system, which combines vestibular signals with visual landmark information to maintain an estimate of the animal's heading, and is a necessary prerequisite for path integration. My students and I have developed computational models which have reproduced a variety of behavioral and neurophysiological observations in these systems, and led to novel predictions.
My second major research area is models of animal learning and their implementation on mobile robots. Current projects include reinforcement learning models of the dopamine reward system, models of classical conditioning based on statistical inference, and a behavioral framework for the Sony AIBO robot.