Modified Project Summary/Abstract Section The cortex is the seat of conscious perception. Our goal is to test how the computations performed by specific cortical cell types lead to specific perceptual states. Here, we will use the mouse visual cortex to determine the cellular basis of contrast perception. Contrast is a fundamental stimulus property represented throughout the visual system and the gain control mechanisms thought to support contrast perception are disrupted in a variety of neuropsychiatric disorders. Thus, identifying the cell types that control contrast perception is a critical step in establishing the neural basis of perception in both health and disease. In our preliminary intracellular recordings from neurons in the visual cortex, we find that increases in contrast drive a transient increase in excitation followed by a prolonged suppression of network excitability. Our hypothesis is that this network suppression is a gain control mechanism that is under the control of local inhibitory interneurons, and is important for determining perceived contrast. Indeed, we find that during performance of a contrast discrimination task, the activation of either parvalbumin- (PV) or somatostatin-expressing (SOM) inhibitory interneurons is sufficient to decrease perceived contrast of stimuli represented by the manipulated hemisphere. To test under what conditions PV and SOM cells contribute to gain control and contrast perception we will use a combination of physiological and psychophysical measurements. In Aim 1, we will use whole-cell and extracellular recordings to determine the stimulus features (i.e. contrast and size) that drive the observed network suppression. Then we will use opto- and chemo- approaches to determine which inhibitory cell types are responsible for driving network suppression under each of these conditions. This will enable dissociation of the functional roles of PV and SOM cells in regulating cortical gain. In Aim 2 we will test the contribution of these cell types to perceived contrast by suppressing their activity during a contrast discrimination task. To establish a link between the circuit mechanisms defined in Aim 1 and the perceptual effects measured in Aim 2, we will incorporate the stimulus conditions that drive network suppression into our contrast discrimination task. This will allow us to compare manipulation of PV and SOM cells on both network suppression and perception across stimulus conditions. Together, these experiments will ascribe cortical computations and cell types to specific perceptual states. This would be a fundamental step in linking the activity of neurons to perception, and the basis for future investigations of the mechanisms of state-dependent sensory processing.