#37 AI Series: Stanford University - S. Sanborn

Join us for an engaging session featuring Sophia Sanborn, an esteemed Assistant Professor at Stanford University! 🌟 She will present on "Virtual Neuroscience with Neural Digital Twins" for approximately 45 minutes. After the talk, make sure to mingle with fellow AI enthusiasts, share ideas, and ask questions while enjoying complimentary drinks and pizza! 🍕🍹

Please note: Doors close by 7:15 PM, so arrive early to secure your spot! It's essential to RSVP here on Meetup to guarantee entry.

Who should attend? This event is designed for anyone intrigued by cutting-edge AI research, particularly students, PhD candidates, academic researchers, and industry professionals specializing in machine learning.

Abstract: The primary aim of systems neuroscience is to decipher how neuron populations represent and transform the natural world. With the increasing scale and complexity of neural datasets, traditional methods of hypothesis generation and neuron characterization become less effective. Enter neural digital twins — deep learning models that predict biological neuron responses. These models simplify the study of large-scale recordings, making them experimentally manageable, allowing for in-silico querying, perturbation, and analysis. Through this approach, we analyze neurons in macaque and mouse visual cortices, uncovering interpretable feature selectivity and revealing previously hidden aspects of the neural code. Moreover, these models facilitate a new avenue of virtual neuroscience, where models of neural systems serve as platforms for automated scientific exploration: searching through stimulus space, mapping selectivity, testing representational hypotheses, and proposing new principles of neural computation.