Join us every Friday at 12:00 p.m. EDT for BME Breaks, Columbia University's weekly webinar series hosted by the Department of Biomedical Engineering. Don't miss the opportunity each week to hear from global leaders in Biomedical Engineering research!

ABOUT THE JUNE 4 WEBINAR

A Less Artificial Intelligence

Despite major advances in artificial intelligence through deep learning methods, computer algorithms remain vastly inferior to mammalian brains: they generalize poorly outside the domain of the data they have been trained on. This results in brittleness (e.g. adversarial attacks) and poor performance in transfer learning, few-shot learning, causal reasoning, and scene understanding, as well as difficultly with lifelong and unsupervised learning — all important hallmarks of human intelligence. We believe current deep learning architectures are severely under-constrained, lacking key model biases found in the brain that are instantiated by the multitude of cell types, pervasive feedback, innately structured connectivity, specific non-linearities, and local learning rules. Therefore, our goal is to learn the brain’s model bias in order to engineer less artificial, and more intelligent, neural networks. Using tour-de-force experimental methods we have been collecting an unprecedented amount of neural responses from the visual cortex, and developed system identification deep learning models that we use to extract principles of functional organization of the brain and learn the brain’s model biases.

ABOUT THE SPEAKER

Andreas Tolias, PhD 
Professor and Brown Foundation Endowed Chair of Neuroscience, Baylor College of Medicine; Founder and Director, Center for Neuroscience and Artificial Intelligence

Dr. Andreas Tolias’ research is focused on understanding how brains give rise to visual intelligence. His lab combines imaging, electrophysiological, molecular and behavioral methods with machine learning approaches to decipher the neocortical circuit principles of perceptual inference. He obtained his B.A. from the University of Cambridge in Natural Sciences and a Ph.D. from the Massachusetts Institute of Technology in Systems and Computational Neuroscience and did postdoctoral training at the Max-Planck Institute for Biological Cybernetics. He has received numerous awards including the NIH Director’s Pioneer Award, the Beckman Foundation Young Investigator Award, the McKnight Foundation Scholar Award, the McKnight Memory and Cognitive Disorders Award and the Michael E. DeBakey Excellence in Research Award. He is a Professor and Brown Foundation Endowed Chair of Neuroscience in the Department of Neuroscience at Baylor College of Medicine and the Founder and Director of the Center for Neuroscience and Artificial Intelligence. Dr. Tolias is also a co-founder of the Neuroscience-Inspired Networks for Artificial Intelligence organization and is leading an international team of scientists and engineers working on the interface between brain research and machine intelligence with the goal of engineering less artificial and more intelligent algorithms. He is also a co-founder of Vathes Inc., Upload AI LLC & BioAvatar LLC