A cortical output channel for perceptual categorization
2026-08-26
Perceptual categorization transforms graded sensory inputs into discrete representations that support decision-making. While the auditory cortex is causally important for this process, a specific circuit location where continuous physical inputs are transformed into discrete decision variables remains unknown. Here, we address this by performing longitudinal two-photon imaging of layer 5 (L5) extratelencephalic (ET) neurons alongside L2/3 and L5 intratelencephalic (IT) populations in mice learning a categorization task. With learning, L5 ET neurons underwent pronounced tuning shifts and developed robust categorical responses, whereas L2/3 and L5 IT neurons did not. This transformation is not merely an intrinsic feature of the cell type but is dynamically controlled by task engagement, emerging only during active decision-making. Using a generalized linear model, we confirmed that categorical selectivity in L5 ET neurons reflects genuine sensory encoding rather than being a by-product of choice-driven activity. These findings identify L5 ET neurons as a cortical output channel to categorize stimuli into a decision variable to guide actions.