Dynamical Principles of Biological and Artificial Neural Networks (Online) (22w5047)


Zachary Kilpatrick (University of Colorado Boulder)

(University of Waterloo)

(University of Alberta)

(York University)


The Banff International Research Station will host the "Dynamical Principles of Biological and Artificial Neural Networks" workshop in Banff from January 9 - 14, 2022.

Artificial intelligence (AI) has recently undergone a revolution, through the use of artificial neural networks, and deep learning. These techniques were developed by simplifying the details of mathematical models of neurons in the brain, and assembling thousands (or in some cases, millions), of these artificial neurons into gigantic networks. This approach has dramatically advanced progress in artificial vision, the control of autonomous vehicles, medical diagnostics, and many other challenging areas of AI research. Spurred on by those successes, neuroscientists have turned to these same artificial neural networks to model the computations performed by real brains: excitingly, the activation patterns of artificial neurons in ANNs often bear strong resemblance to those observed in the real brain.

Despite these recent successes at the neuroscience-AI intersection, there are many deep mathematical questions still to answer: 1) given their enormous size and complexity, how can we understand the operations performed by a fully-trained deep learning algorithm, or (analogously) an animal’s brain?; 2) how does the dynamics of activations within biological and artificial neural networks evolve and interact over different timescales (seconds, hours, days, etc.)?; 3) how does randomness in the environment and/or within the neuronal networks themselves contribute to the function of these networks?; 4) what features most enhance (or degrade) the computational power of neuronal networks?; 5) how do brains and AI systems reason in the face of uncertainty? This upcoming BIRS workshop will bring together AI researchers, neuroscientists, and mathematicians, to answer these questions. In so doing, it aims to advance all three of those disciplines, leading to better AI algorithms, and potentially better diagnosis and treatment of those with brain disorders.

The Banff International Research Station for Mathematical Innovation and Discovery (BIRS) is a collaborative Canada-US-Mexico venture that provides an environment for creative interaction as well as the exchange of ideas, knowledge, and methods within the Mathematical Sciences, with related disciplines and with industry. The research station is located at The Banff Centre in Alberta and is supported by Canada's Natural Science and Engineering Research Council (NSERC), the U.S. National Science Foundation (NSF), Alberta's Advanced Education and Technology, and Mexico's Consejo Nacional de Ciencia y Tecnología (CONACYT).