The Neuroscience of Artificial Neural Networks: From Inspiration to Analysis
NEURO 101GG
Subject & Catalog Number
Course Information
Description
How “neural” are artificial neural networks? Drawing on foundational and modern research, this course explores the conceptual and mathematical parallels between biological and artificial neurons, tracing their intertwined histories from early perceptrons to today’s large-scale deep networks. We will investigate fundamental similarities and key divergences—such as with neural selectivities, large scale organization, and learning rules—and discuss how neuroscience methods are being used to interrogate the internal structure of deep vision and language models today. By examining seminal and contemporary literature, students will build an intuition for how brain-inspired principles have shaped modern AI, the role of neuroscience in AI development today, and how neuroscientific techniques can offer unique insights into the behavior and organization of complex computational systems.
Class Notes
Enrollment for this course will be via lottery with preference given to junior Neuroscience concentrators. Lottery instructions, deadlines, and a link to the google form can be found here: https://www.mcb.harvard.edu/undergraduate/neuroscience/neuro-courses/?course-button=tutorials
Available for Harvard Cross Registration
