Project Summary: Hallucinations are common in clinical and nonclinical groups, can be difficult to treat, and often predict worsening functionality. The poor efficacy and severe side effects of current treatments are in part a consequence of our immature understanding of the mechanisms that cause hallucinations. Excess striatal dopamine release has been causally implicated in the development and severity of hallucinations but the precise circuits and cognitive processes that link this neurochemical alteration to false perception remain unclear. Evidence from the basic neuroscience literature has inspired competing theories about how excess striatal dopamine drives hallucinations. Specifically, reward and perceptual hypotheses of hallucinations have emerged, but they have yet to be directly tested in a falsifiable framework. Identifying which of these hypothesized mechanisms drives hallucinations is critically important given that reward and perceptual learning are facilitated by distinct dopaminergic basal circuits each of which may provide a separate treatment target. We have developed a mathematical framework that formalizes these hypotheses with biologically grounded computational models and generated falsifiable predictions about how alterations in either perceptual or reward learning could drive hallucinations. Here, we will rigorously test the neural and behavioral predictions of these models using a novel fMRI-compatible auditory signal-detection task and a validated proxy measure for midbrain dopamine function. In Aim 1, we will evaluate participant perceptual and reward learning and the relationship with hallucination proneness. In Aim 2, we will identify the neural circuits that support reward and perceptual learning during the task. In Aim 3, we will use a validated proxy measure of dopamine function to dissociate the specific subcircuits driving alterations in learning. Overall, the proposed study aims to bridge the explanatory gap between our understanding of the neurochemistry and phenomenology of hallucinations. Critically, this could promote the identification of individualized treatment targets that are not only more effective but have more limited side effects. This proposal will also support my training in state-of-the-art computational modeling and neuroimaging approaches and promote my development as an independent researcher in the field of computational psychiatry.