Research

We investigate how neural circuits in the brain perform biologically relevant computations, such as: maintenance of short term memory, sensory inference, spatial computation, and generation of sequential neural activity. Several areas of current interest are listed below.

 
 

Maintenance of short-term memory in the brain

How does the brain maintain short-term memory in sustained neural and synaptic activity? What are the constraints posed by neural hardware and by noise? We are interested especially in representations of continuous variables, such as orientation, eye position, location in space, or accumulated evidence, since this type of representation is particularly sensitive to noise.


Spatial representation and computation in the mammalian brain

We investigate theoretically mechanisms and principles that underlie spatial representation in the brain. We are particularly interested in the emergence and dynamics of grid cell activity in the entorhinal cortex, in the interaction of grid cells with other spatially selective cells in the hippocampal formation, and in the functional role of grid cells in the brain.


Fixational eye motion

Fixational drift in the primate and human visual system is a random, smooth motion of the eyes that occurs between microsaccades, during fixation. On the scales of foveal vision the range of fixational drift is surprisingly large to the extent that, naively, it should blur the image perceived by the brain - in similarity to the blurring of a photograph due to shaking of the hands. However, our visual system achieves exquisite acuity. We investigate the consequences of fixational drift for perception, and its mechanistic origins.


Synaptic plasticity in recurrent neural networks

We investigate how local plasticity rules, acting on individual synapses in recurrent neural networks, give rise to large-scale, global structures that can support various forms of computation.


 
 

Updated: April 2022