Multivariate Representation of Sustained Visual Content in a No-Report Paradigm
Gal Vishne, Edden M. Gerber, Leon Y. Deouell, Hebrew University of Jerusalem, Israel; Robert T. Knight, University of California Berkeley, United States
Contributed Talks 2 Lecture
Grand Ballroom A-C
Fri, 26 Aug, 16:50 - 17:10 Pacific Time (UTC -8)
Vision research has thoroughly characterized the information content of transient neural responses to stimuli onset. Yet, visual experience contains many stationary moments when we still maintain information about our environment, highlighting the need to understand this aspect of perception. To investigate visual representation over time, we utilize intracranial recordings from patients viewing images of variable durations and categories (focusing on no-response trials), an approach promoted recently to arbitrate between two prominent scientific theories of consciousness (Melloni et al., 2021). We previously showed single-electrode responses in category selective regions are dramatically attenuated after the onset (Gerber et al., 2017). Here, we examine category and single-item representation using multivariate state-space analyses, decoding, and representational similarity analysis. First, we show sustained and stable visual representation in ventral-temporal and occipital cortex, despite the reduction in activity. Second, we show transient representation of visual content in prefrontal cortex, even though no report was required. This suggests that ongoing content is maintained in sensory regions, and prefrontal cortex responds to changes regardless of the task. Importantly, as predictions of both consciousness theories bear out, it questions whether this approach can arbitrate between them.