Organization of neural coding and plasticity in L2/3 of mouse S1 cortex - Summary Non-topographic, intermixed representations (salt-and-pepper maps) of sensory information are common in cerebral cortex, but how neural coding and plasticity are organized within them is unclear. We propose that salt-and-pepper maps contain distinct pyramidal (PYR) subnetworks with differential roles in coding stability and flexibility (including learning and attentional modulation). To test this, we study the whisker map in layer 2/3 of mouse somatosensory cortex (S1), where PYR cells tuned for the columnar whisker (CW) and for non- columnar (non-CW) whiskers are intermixed in each column. We recently discovered that non-CW tuned cells show marked tuning instability across days, while CW-tuned cells have stable tuning. This reveals that the L2/3 salt-and-pepper map has two components: a stable columnar map of CW-tuned cells, intermixed with non-CW tuned cells that are unstably tuned and have little columnar topography. We propose that CW- and non-CW tuned cells are distinct PYR subcircuits with different roles in coding and plasticity. This is a novel model of S1 circuit function. We predict that the CW network provides coding stability, while non-CW cells are the primary site for plasticity and learning. Based on preliminary data, we hypothesize that tuning instability in non-CW cells is internally driven, and acts to sample novel sensory codes which may then be stabilized by experience or reward. This is a novel hypothesis for how sensory maps balance stability and plasticity—by segregating these functions in different subcircuits. In Aim 1, we use longitudinal 2-photon calcium imaging to understand the nature and origins of tuning instability, and to test whether experience or reinforcement stabilizes whisker tuning. In Aim 2, we evaluate whether CW and non-CW networks represent distinct functional networks with different sensory coding and plasticity properties. We test our central hypothesis that non-CW cells are the primary locus of sensory plasticity and learning within the map. Aim 3 asks how attention modulates neural coding within intermixed maps. We developed a selective attention task in which mice use history-dependent cues to guide attention to a specific whisker to improve detection performance. Mice show robust spatial attention to cued whiskers. Attention lasts ~10 sec and is driven by recent pairing of whisker stimuli with reward. Preliminary data show that attention enhances whisker-evoked activity of PYR cells encoding the attended whisker in S1. This establishes S1 as a powerful site to study cortical mechanisms of attention. We will use 2-photon imaging and Neuropixels recording to study how attention modulates sensory coding in S1, including measuring the size and CW- or non-CW network specificity of the attentional spotlight. In a major effort, we use imaging and optogenetics to identify the control circuits for attention in S1, with initial focus on VIP interneurons. Together, these studies will reveal how plasticity and attentional modulation are organized within a canonical salt-and-pepper map.