Neuronal Features of Visual Attention in the Mouse Superior Colliculus Depend on Learned Behavioral Relevance

L Lupeng Wang C Christian Quaia K Kerry Elliott K Kara K. Cover R Richard J. Krauzlis

Abstract

Visual selective attention is not wired into the brain fully formed and immutable but is acquired and refined as animals learn by interacting with their environment. Here we investigated this process by studying neuronal activity in the superior colliculus (SC) of male and female mice that learned different meanings of the same visual stimuli. We recorded spiking activity of neurons across the superficial and deep SC layers in three cohorts of mice, each trained using a tailored set of task rules that attributed different levels of behavioral relevance to the visual stimuli. Experimental conditions across the three tasks were carefully matched for visual stimulation and task engagement. We found that several markers of attention-related activity depended on the level of learned behavioral relevance, with the strongest effects in deep SC. First, for activity evoked by cue onset, visual responses in superficial SC were larger in animals that exhibited higher learned relevance. In deep SC, the presence of visual onset responses depended entirely on the learned behavioral relevance. Second, traditional attention-related modulation, defined as enhanced steady-state activity for the visual stimulus at the cued location, was also stronger with higher learned relevance, and this effect was limited to deep SC. Finally, the response to the visual change at the cued location was virtually absent when learned behavioral relevance was low but was prominent after visual training. Together, these results reveal that several fundamental features of attention-related modulation depend on the behavioral relevance learned through task-specific training, including aspects of stimulus-driven attention.

Article Details

Volume / Issue Vol. 46, Issue 5
Published February 04, 2026
Pages e1187252025
ISSN 0270-6474
Publisher Society for Neuroscience

Journal Info

Journal of Neuroscience

Society for Neuroscience

ISSN: 0270-6474 Life Sciences

Authors (5)

L

Lupeng Wang

C

Christian Quaia

K

Kerry Elliott

K

Kara K. Cover

R

Richard J. Krauzlis