Ila R. Fiete

Scholar: 2010

Awarded Institution
Assistant Professor
The University of Texas at Austin
Center for Learning and Memory


Research Interests

Neural Coding and Dynamics

The power of neural computation is all the more remarkable given that it arises in part through self-organization and is carried out in a milieu of noisy processes. My group uses computational and theoretical tools to better understand the dynamical and coding principles that underlie such computation.

Our efforts on the dynamics front aim to answer the `how' question: How do recorded neural responses arise out of collective dynamics and network plasticity? Our work on neural coding is focused on the `what' and `why' questions: What information is encoded in neural activity? And why is information encoded the way it is -- more concretely, what are the properties of the coding scheme that make it suitable for encoding that variable?

Of particular interest is to understand how the cortex can perform precise computations given the level of noise measured in cortical neurons and synapses. We hypothesize that neural codes are selected to have strong performance on error self-checking and correction, in the spirit of "good" error control codes defined by Shannon. We are exploring this hypothesis by considering specific cortical codes. Our aim is to gain novel insights into the structure and priorities of neural coding and to generate predictions about functional connectivity within the coding region and between the coding region and its downstream decoders.