How category frameworks shape the pre-activation of attentional templates: Evidence from high-resolution SSVEPs

Abstract

During visual search, when targets are defined by specific categories, category-based attentional templates (CATs) enhance search efficiency by being pre-activated and modulated by their underlying category frameworks, thereby facilitating the detection of target objects. Previous studies employing the Rapid Serial Probe Presentation (RSPP) paradigm (20 Hz sampling rate) lacked the temporal resolution to continuously track CAT pre-activation dynamics. To address this, we used steady-state visual evoked potentials (SSVEPs) to capture real-time dynamics of CAT pre-activation, improving temporal resolution by 50 times (from 50 ms to 1 ms), enabling the detection of previously inaccessible transient neural processes. SSVEPs and multivariate pattern analysis (MVPA) revealed distinct pre-activation time courses for different CAT frameworks. Specifically, prototype-based CATs (involving perceptual similarity) were pre-activated earliest in posterior visual regions (1808 ms pre-target), followed by semantic-based CATs (involving experience and learning) in anterior temporal networks (1460 ms pre-target), and strategy-based CATs (involving learning) engaging prefrontal control systems last (359 ms pre-target). By uncovering the temporal hierarchy of CAT pre-activation across different cortical systems, our findings refine predictive coding theory, demonstrating how category frameworks shape the temporal dynamics of anticipatory attention and guide CAT pre-activation during visual search.