PROJECT SUMMARY / ABSTRACT Persistent activity in neural circuits supports a variety of brain functions from motor control to navigation to perceptual decision-making. Correlational studies show significant variation in persistent activity patterns during different behaviors, suggesting that individual circuits perform flexible computations that depend on the context of ongoing brain activity and motor functioning. However, establishing the causal significance of this variability is difficult due to technical limitations in existing tools for precisely manipulating circuit dynamics. A tractable system for overcoming this challenge is the zebrafish, a vertebrate model organism with an optically accessible brain that allows simultaneous calcium imaging and optogenetic stimulation with laser microscopy. Research in our lab focuses on the zebrafish oculomotor integrator, a hindbrain circuit involved in adaptive control of gaze position. This circuit generates persistent activity that directly drives easily quantified motor behavior. In prior work, our lab found that different patterns of integrator activity are associated with distinct types of eye movements, but it is unclear how these context-dependent dynamics contribute to oculomotor control and how they relate to the development of sophisticated behavior. In the proposed research, I will conduct simultaneous two-photon imaging and optogenetic stimulation during visuomotor behavior to determine how different patterns of integrator dynamics contribute to different types of eye movements. First, I will collect a comprehensive longitudinal dataset of brain-wide neural activity in the larval and juvenile zebrafish during a broad range of oculomotor behaviors, testing if indeed the number of persistent patterns in the integrator expands with the behavioral repertoire. Then, I will perform real-time closed-loop stimulation of the integrator network at single- cell resolution, steering circuit dynamics along specific patterns of activity to test their causal impact on motor outputs. This research will improve our understanding of flexible control by memory circuits and establish new paradigms for precise manipulation of network dynamics.