Event Details:
Continue the conversation: Join the speaker for a complimentary dinner in the Theory Center (second floor of the neurosciences building) after the seminar
Toward a Framework for Investigating Neural Computation with Empirically Detailed Neuron Models
Abstract
Biological neural networks within the brain are energy efficient and are the basis of intelligence, yet traditional neuron and network models often abstract away the intricate dendrite structures and their rich subcellular complexity observed in real neurons. Although artificial neural networks (ANNs) are foundational to incredibly high-performing artificial intelligence applications, they rely on simplified, point-like neurons that neglect the nonlinear processing inherent to neuronal dendritic structures. Here, we investigate the computational capabilities of individual neurons, aiming to refine our expectations of the overall computational power of biological networks. First, we model a single neuron as a multilayer network that integrates key dendritic structural properties and nonlinearities. We demonstrate that optimizing these single neuron models to perform nontrivial computer vision tasks (e.g., MNIST and Fashion-MNIST) reveals how subcellular properties synergize to enhance individual neuron computation. Furthermore, to expand our investigation of neural computation to dynamic neuron models with biophysical mechanisms, we introduce a gradient-based optimization technique leveraging differentiable ODE solvers. This addresses the inefficiencies of traditional gradient-free optimization methods for high-dimensional, dynamic neuron models by enabling high-dimensional parameter fitting, paving the way for biophysically realistic neuron and network simulations. Our work establishes an optimization-based framework for future investigation of the level of biological detail needed to effectively study biological network computation.
Ilenna Jones
Harvard University
Ilenna Jones is a Research Fellow at the Kempner Institute for the Study of Natural and Artificial Intelligence at Harvard. Her research interests include neuronal biophysics and computation, model optimization, and neuroscience for AI. She received her B.A. in Neuroscience in 2015 from Dartmouth College. In 2023 she received her PhD in Neuroscience at the University of Pennsylvania in Konrad Kording’s laboratory. Ilenna began her position as a Research Fellow in the Kempner Institute in 2023. There she continues her work in the neuro-AI space to investigate how subcellular neuronal properties enable single-neuron and network computation using principles from optimization, deep learning, and biophysics.
Hosted by Alice Tor (see profile below)
About the Mind, Brain, Computation, and Technology (MBCT) Seminar Series
The Stanford Center for Mind, Brain, Computation and Technology (MBCT) Seminars explore ways in which computational and technical approaches are being used to advance the frontiers of neuroscience.
The series features speakers from other institutions, Stanford faculty, and senior training program trainees. Seminars occur about every other week, and are held at 4:00 pm on Mondays at the Cynthia Fry Gunn Rotunda - Stanford Neurosciences E-241.
Questions? Contact neuroscience@stanford.edu
Sign up to hear about all our upcoming events
Related Event
