Our goal is to understand the biological basis of cognition. After many experimental studies elucidating the rules of associative synaptic modification, almost all of our research uses the tools of theoretical neuroscience to reach our goal.
Using computational simulations of the hippocampus and associated cortical regions, the laboratory has developed a model of the hippocampus that is biologically realistic and that can perform many cognitive tasks that depend upon an intact hippocampus. These tasks include transitive inference, transverse patterning, trace conditioning, and maze learning. The model is currently being extended to include interactions with nucleus accumbens, the amygdala, and the orbital prefrontal cortex.
We are also studying the physical constraints on communication and computation in the brain. Optimization results that combine information-theoretic perspectives with energy constraints explain several measured properties of neurons.
The Big Picture
Finally, there is a software engineering project aimed at producing a verified but flexible approach to large scale neural network simulations. This integrated package is being developed to simulate the biology of neurons and synapses and, as well, to simulate the interaction of an organism with the environment. For more details, please visit the Neurojet Website.