Modulation of sensory processing

Neurons in sensory cortex do not respond to the stimulus alone: their activity is also shaped by behavioral context and the internal state of the animal, including attention, arousal, task demands, decision-related feedback, and movement. Disentangling these influences from stimulus-driven activity is difficult, because they overlap in time and are often only indirectly measured. We use statistical models of neural populations, including latent-variable models that infer unobserved shared factors, to separate stimulus-driven responses from state- and behavior-related modulation. We collaborate with the Nienborg lab (NEI) and Huk lab (UCLA) to apply these latent-variable models to recordings in awake primates, asking how much of visual cortical activity is explained by spontaneous movements, attention, eye movements, and top-down feedback, and where in the visual hierarchy these effects arise.

Relevant publications and presentations