Visual cortical dynamics supporting predictable attentional capture

Visual behavior depends on the ability to prioritize relevant sensory information while filtering out distractions. Predictable sensory contexts enable more efficient behavior by altering sensory processing. Using laminar neurophysiology in macaque visual cortex measuring population spiking during a feature-based pop-out visual search task, we examined how predictable visual routines influence cortical columnar processing of sensory information. By manipulating predictability through attentional priming, we found improved behavioral performance with predictable stimulus arrays. These behavioral changes were supported by earlier neuronal attentional target selection driven by reduced variability in the sensory response to the predictable target stimulus and more homogeneous feedforward processing dynamics across the cortical column. More effective distractor suppression was driven by adapted feedforward processing of the more frequent distractor stimulus. Taken together, these changes implicate independent mechanisms for target enhancement and distractor suppression. Our results highlight how predictions from experience alter visual columnar processing to optimize attentional selection through streamlining feedforward signaling. HighlightsO_LIBoth target enhancement and distractor suppression are independently observable in visual cortex with different profiles of modulation as a function of predictability. C_LIO_LIPredictions produce stronger, more reliable feedforward responses to attention-capturing targets across the visual cortical layers. C_LIO_LIBetter distractor suppression with stimulus predictability is driven by adapted sensory representations, not top-down inhibition. C_LI SignificanceHow do predictions help determine what catches our attention? It is well known that when primates repeatedly perform visuomotor tasks, they become more proficient. Using neural recordings across the layers of the visual cortex, Groot et al. demonstrate that more efficient feedforward processing of sensory information contributes to improved visual behavior under predictable sensory conditions. These findings highlight the crucial role of predictions in proactively shaping sensory representations to facilitate efficient attentive behavior.