Huntret · Context-Dependent Retinal Computations During Innate Visual Behavior in Mice

Understanding how perceptual input is transformed by neurons into behavioral responses is a central question in neuroscience. Recent work indicates that, for ethologically relevant stimuli—like prey pursuit and predator avoidance—neural correlates of behavior already emerge within retinal circuits, challenging the view of the retina as a mere low-level feature detector.This project will test the hypothesis that retinal ganglion cells (RGC) contribute to these behaviors depending on the visual context. Natural environments are visually complex, yet simplified stimuli are common in vision research, probably obscuring context-dependent retinal computations. To overcome this, I will develop an integrated pipeline that combines naturalistic behavior, retinal electrophysiology, and computational modeling. First, I will quantify how complex backgrounds modulate mice hunting performance in an immersive arena. Then, using mice-perspective videos from behavioral trails, I will stimulate retinal explants on multi-electrode arrays to record RGC responses, the retinal output. Finally, I will train convolutional neural networks on this data to create predictive models describing the RGC computations, which will be finally validated in the behavioral arena. This approach will reveal how background modulates retinal output to drive prey capture, closing the loop from neural responses to behavior. Ultimately, this project moves beyond reductionist approaches by using naturalistic stimuli to study the system as a whole, probing its responses to the very inputs that shaped its evolution.My expertise in retinal physiology and behavior perfectly complements the host's proficiency in electrophysiology and modeling. This synergy uniquely positions me to successfully execute this interdisciplinary project. After completion, this fellowship will equip me with the necessary skills I need to create an interdisciplinary research team in the emerging field of neuroscience of natural behavior