Watch online : 【Popular Science Lectures (Online)】Does the brain-inspired deep learning help us learn the brain?—A neurophysiological perspective
Following the groundbreaking achievements made by deep learning methods including AlphaFold and RoseTTAFold for predicting protein structures—a fundamental, difficult challenge in biology—this apparent coming-of-age of artificial intelligence (AI) again electrified the world. The research paradigm of artificial neural network rose from the crossovers of Cognitive Neuroscience and Machine Learning, rooted in the inspirations provided by brain studies. Now, the question is being asked the other way around: as the most ‘intelligent’ models the human has ever built, could AI teach us anything about the essence or principles of how the brain supports complex behavior? In this talk, I will attempt to initiate a quest for such possibilities from a general standpoint related to Neurophysiology, a time-honored discipline that aims to understand how the nervous system works.