Doris’ little corner of the internet

I’m currently a postdoctoral scholar at Mila and at Université de Montréal, working with Dr. Shahab Bakhtiari. I was previously a postdoc in Seattle at the AI Institute in Dynamic Systems working with Nathan Kutz and Steve Brunton. I consider myself to be an applied mathematician who enjoys thinking a lot about dynamical systems, statistical tools for data modeling (including neural networks and LLM’s!), computational neuroscience, the link between neuroscience and AI models (neuroAI), and generally about intelligent systems.

I have a general interest in data-driven modeling of complex dynamical systems in science, as well as developing novel machine learning/AI tools to better understand datasets that describe these systems. I recently developed a dynamic SINDy, a machine learning tool for system identification in noisy, nonlinear, and non-autonomous dynamical systems.

I have been focusing on applications to neuroscience, where modeling dynamics can be very challenging… Since starting my Ph.D, and even before that when I was an undergrad, I’ve been fascinated with understanding computation in simple and complex nervous systems like our brain. Nowadays, I like to think about computation through neural population dynamics.

I did my Ph.D. at University of Washington under the guidance of Eric Shea-Brown and Stefan Mihalas, while being a visiting student at Allen Institute for Brain Science. My doctoral work looks at computational models of the visual cortex that can explain its adaptability to switching contexts, specifically when animals are static versus running. I was/am also interested in biologically-inspired deep learning architectures, such as those that can switch between contexts just like the visual cortex.