|
|
 
University of Pennsylvania
David Rittenhouse Laboratories
209 South 33rd Street
Philadelphia, PA 19104
Office: DRL 2N3A
Office phone: 215.573.0982
Lab: DRL 2N3A
E-mail: vijay@physics.upenn.edu
Research Interests
What determines the algorithms used by neurons to compute and transmit
information? My work aims to develop a theory of the design of neural
circuitry. In the sensory periphery this design must be dominated by
three factors: (a) where information resides in natural stimuli, (b)
the constraints inherent in biological information processing (energy
costs, space consumption, temporal limitations), and (c) noise. My recent
work explores the theory of energy efficient information processing,
and shows that the codes implemented by retinal ganglion cells can optimize
bits transmitted per energy subject to noise. We are currently exploring
the generality of this principle of energy efficiency under a variety
of stimulus conditions and for a variety of cell types in the retina.
This also leads to a second question: Why is there more than one kind
of cell? For example, since a general computation can be carried out
by NAND gates, why are there so many different kinds of ganglion cells
in a retina? Presumably, in order to extract behaviourally relevant
information from natural scenes in a timely manner, subject to biological
constraints such as noise and energy/space limitations, many different
types of cells or circuit elements must be present. With this in mind
I am exploring how the structure of natural scenes influences the design
of retinal cells. This work is being carried out in collaboration with
Peter Sterling's lab in the department of neuroscience.
Graduate Groups
|
 |
