AI x Science Seminar: Eva Dyer
Building unified models of neural data across tasks, modalities, & species
Amy Gutmann Hall, Room 414
Neural activity is complex, dynamic, nonlinear, and high-dimensional, and the datasets we collect from the brain are equally varied. This richness makes analysis difficult, and models trained on narrow conditions rarely generalize. Rather than seeking to simplify the problem, we believe progress requires embracing this complexity. In this talk, I will describe our efforts to build models for neural data that integrate many datasets into a unified map of brain activity, which can then be adapted to diverse downstream tasks. I will highlight results from our group showing how scaling provides concrete benefits, including transfer across brain regions, individuals, and even species. Looking ahead, we believe that unified models for neural data have the potential to transform how we chart brain function, uncover principles of neural computation, and reimagine interventions for mental health.
Sponsored by Penn AI, Innovation in Data Engineering and Science (IDEAS), and the Data Driven Discovery Initiative (DDDI).