PROJECT SUMMARY Sensorimotor transformations are mediated by premotor brain networks where individual neurons represent sensory, cognitive, and movement-related information. In the superior colliculus (SC), a central hub for producing visually-guided saccadic eye movements, many neurons are active during all three stages, emitting transient, high-frequency bursts of spikes during the sensory and motor events and exhibiting persistent, lower firing rate activity in-between the bursts. The mixed-selectivity to multiple dimensions of information exemplifies a potentially efficient mode of information representation by the nervous system, but it also raises crucial questions: What features differentiate the two bursts, and how does a decoder know precisely when to initiate a movement if its inputs are active at times when a movement is not desired (e.g., in response to sensory stimulation)? What information is encoded in the low-frequency activity, and how is it modulated during different cognitive demands? We reason that the answers to these questions lie not in the activity of individual neurons but rather across the population of active neurons and in the temporal dynamics. Specific Aim 1 tests various neural mechanisms of movement initiation by quantifying features that differentiate the sensory and motor bursts across the SC population. Specific Aim 2 focuses on the low-frequency activity that intervenes between the two bursts. We will use a dynamical systems approach to characterize how the neural trajectory evolves during sensorimotor transformation and how it differs for tasks with different cognitive loads. The ability to discriminate neural trajectories according to task demands indicates a potential mechanism by which different dimensions of information can be multiplexed into the same population of neurons.