Friday, April 11
SEC LL2.221 or Zoom (Passcode: 988031)
11:00am - 12:00pm

"Constrained Models of Neural Dynamics for Generalization and Insights"
Shreya Saxena, Assistant Professor of Biomedical Engineering, Yale University

Abstract: Our ability to record large-scale neural and behavioral data has substantially improved in the last decade. However, the inference of quantitative dynamical models for cognition and motor control remains challenging due to their under-constrained nature. Here, we incorporate constraints from anatomy and physiology to tame machine learning models of neural activity and behavior.

How does the motor cortex achieve generalizable and purposeful movements from the complex, nonlinear musculoskeletal system? I will introduce a deep reinforcement learning framework that trains recurrent neural network controllers to generate purposeful movements in anatomically accurate macaque and mouse musculoskeletal models. This framework mirrors biological neural strategies and aids in predicting and analyzing novel movements. In the second part, I will present switching recurrent neural networks to detect cognitively and behaviorally relevant states in neural data, capturing underlying nonlinear dynamics. Finally, I will discuss ongoing work on integrating region-specific constraints in models of the cortico-basal ganglia-thalamic loop during timing tasks to gain insights into pathway-specific computations. Through these projects, we show that a constraints-based modeling approach allows us to predictively understand the relationship between neural activity and behavior.
 

Bio: Shreya Saxena is broadly interested in the neural control of complex, coordinated behavior. She is currently an Assistant Professor of Biomedical Engineering and an Investigator at the Center for Neurocomputation and Machine Intelligence at the Wu Tsai Institute at Yale University. Before this, she was at the Department of Electrical and Computer Engineering at the University of Florida as an Assistant Professor from October 2020 to June 2023. During Shreya’s postdoctoral research at the Center for Theoretical Neuroscience at Columbia University’s Zuckerman Mind Brain Behavior Institute, she developed machine learning methods for interpretable modeling of neural and behavioral data. Her PhD in the Department of Electrical Engineering and Computer Science at the Massachusetts Institute of Technology (MIT) dealt with performance limitations in sensorimotor control. Shreya received an M.S. in Biomedical Engineering from Johns Hopkins University, and a B.S. in Mechanical Engineering from the École Polytechnique Fédérale de Lausanne (EPFL). She is honored to have been selected as a 2025 Alfred P. Sloan Fellow, and a Rising Star in both Electrical Engineering (2019) and Biomedical Engineering (2018).