Home Department: Symbolic Systems
Mentor: Thomas Clandinin, Neurobiology

“Investigating the Role of Recurrent Connectivity in Visual Processing”

Feedback and recurrent connectivity are ubiquitous but poorly understood features of the nervous system, and they underlie state-of-the-art machine learning models for speech recognition, robotics, and other tasks. As such, demystifying the functional role of recurrence could greatly contribute to our understanding of the brain and help us engineer better AI systems. To that end, Mohammed will be using computational models to investigate how recurrent connections contribute to motion detection in the fruit fly visual system.