Semantic memory and abstraction of knowledge from experience are thought to involve communication between the hippocampus and neocortex. The hippocampus encodes novel episodes using sparse and orthogonal neural representations, which form and stabilize within ~10 minutes in a novel environment. Recent studies have shown that several neocortical areas, including the retrosplenial cortex (RSC), primary somatosensory cortex (S1), and primary visual cortex (V1) contain sparse neural coding correlated to spatial location in one-dimensional environments. This spatial coding depends on an intact hippocampus, at the time of learning. We recently found that superficial RSC pyramidal cells, like hippocampal neurons, rapidly formed spatial representations in novel (virtual) environments, but only in the presence of stable visual cues. This suggests that a correspondence between bottom-up sensory and spatial (top down) signals is required to form spatially correlated activity in neocortex. Therefore, we asked whether different cortical areas would equally form position-correlated activity in the presence of visual cues alone. We used 2-photon imaging to simultaneously study activity in RSC, V1, and S1. We predicted that V1 and RSC would be able to form an association between visual cues and spatial signals, and thus express consistent spatial activity, while S1 would not. We found that indeed the activity of S1 neurons could not decode position in a visual environment as well as simultaneously recorded V1 and RSC neurons. We conclude that the expression of position-linked memory indexes in neocortex requires a consistent correspondence between unique modality-specific cues and spatial information from hippocampus during learning.