Abstract
Cortical neurons projecting to the same target area may form specialized population codes to transmit information, but whether and how they do so remains unclear. We used calcium imaging in mouse posterior parietal cortex, retrograde labeling and statistical multivariate models to address this question during a delayed match-to-sample task in virtual reality. We found that neurons projecting to the same area have elevated pairwise activity correlations. These correlations are structured as information-limiting and information-enhancing motifs that shape interaction networks and collectively enhance information about the mouse's choice beyond what is contributed by pairwise interactions. This network structure is unique to subpopulations that project to the same target and was not observed in surrounding neural populations with unidentified projections. Furthermore, this structure is only present when mice make correct, but not incorrect, behavioral choices. Therefore, cortical neurons comprising an output pathway form a population code with a unique correlation structure that enhances population-level information to guide accurate behavior.