INTRINSIC POPULATION DYNAMICS ARE A NEURONAL SUBSTRATE FOR VISUAL SELECTIVE ATTENTION

Perception results from a dynamic interplay between the feedforward processing of sensory stimuli and intrinsic neural activity, which is often dismissed as noise. To tailor perceptual processes to the organism’s current needs on a continuous, moment-to-moment basis, intrinsic dynamics – rather than just being noise – have been suggested to reflect prior expectations, task demands, and attentional focus. Here, we identify an intrinsic, collective neural phenomenon in the retinotopically organized superior colliculus (SC), a midbrain hub integrating bottom-up visual input with top-down signals to guide visuospatial attention and saliency mapping. We show that these intrinsic dynamics organize cell activity into structured “blob-like” spatiotemporal assemblies that are temporally and topographically localized and that rival sensory-evoked responses in strength. These dynamics fluctuate with learning stage and attentional engagement, enhance visual responses in a manner that mirrors behavior, and predict trial-by-trial context and performance in a visual attention task. Although independent of sensory input or overt behavior, these assemblies can be recruited during attentive states by visual stimuli, generating a “dynamic saliency map” that aligns with reaction times and behavioral outcomes. Modeling indicates that tunable blob-like assemblies can arise from local excitatory–inhibitory interactions within the SC, and enhance sensory-evoked responses in line with observations. Together, our results show that intrinsic population dynamics are necessary for shaping a dynamic saliency map and driving goal-directed actions during a spatial attention task.