I work on the theory and modeling of neural systems where the collective behavior is rich (emergent), but where underlying features such as single-neuron properties and local connectivity are constrained by experiment. My aims are to help (1) Elucidate the essential dynamical principles underlying emergent motor and sensorimotor function, (2) Understand principles of the encoding and decoding of neural information based on system function, and (3) Drive fruitful interactions between theory and experiment by generating non-trivial predictions for neural organization, activity, and synaptic plasticity. Ongoing and recent projects investigate:

Precision and robustness of the dynamics of spatial path integration in the rodent entorhinal cortex

Principles and capacity of position encoding in rodent entorhinal cortex

Decoding or readout by hippocampus of position-related information from entorhinal cortex

Emergence of ultrasparse representations in neural networks, and of ultrasparse sequences in songbird vocal premotor regions

General synaptic rules for goal-directed learning in recurrent networks of conductance-based spiking neurons

Goal-directed sensorimotor song learning in songbirds

The role of sparse codes in learning in feedforward neural networks