MOUSE V1 ENCODES SCENE CONTEXT RATHER THAN STIMULUS SALIENCY

Visual behavior depends on both saliency-driven attention and context-dependent perception, yet how early visual circuits differentially represent these signals remains unclear. We addressed this question by examining how neurons in the superior colliculus (SC) and primary visual cortex (V1) respond to salient flash gratings (SFGs) presented at different spatial locations relative to their classical receptive fields (CRFs). Robust SFG responses were observed in only a subset of V1 neurons, whereas most SC neurons responded reliably, reflecting the greater diversity of receptive field structures in V1. Among V1 neurons responsive to both salient stimuli and background, maximal responses occurred when the target was displaced approximately 10 degrees from the CRF center, unlike SC neurons, whose responses peaked when the target is at the CRF center. Furthermore, V1 responses were strongly modulated by background orientation even when saliency strength was constant: responses to SFGs on preferred versus orthogonal backgrounds were anti-correlated. By contrast, SC neurons encoded saliency largely independent of orientation preference. These findings indicate that SC provides a robust, preference-independent saliency signal, whereas V1 preferentially represents scene context. This functional dissociation highlights complementary computational roles of early visual pathways in supporting attention and perception.