- Main
Mechanistic challenges in achieving precise top-down attentional modulations in the visual system
- Park, Sunyoung
- Advisor(s): Serences, John
Abstract
Voluntary selective attention is a fundamental process in visual perception, enabling the prioritization of behaviorally relevant information from a complex sensory environment. While prefrontal cortex (PFC) has been implicated as the source of this top-down modulatory signal, how this is achieved is unclear given that PFC neurons have coarse and high-dimensional selectivity. Thus, this dissertation investigates how prefrontal areas with coarse and high-dimensional selectivity for spatial locations and visual features modulate sensory processing with high precision. Chapter 1 explores the limits in the precision of top-down spatial attention, revealing that feedback signals from frontal and parietal regions are less spatially precise compared to the sensory responses in early visual areas, such as V1. Chapter 2 uses spiking neural network models to examine how feedback signals, despite their coarse and high-dimensional nature, can lead to precise modulations of early visual processing. These simulations highlight the importance of random projections between control and sensory networks in producing targeted attentional modulations and controlling disruptive spurious signals. Chapter 3 further investigates the impact of structured versus random feedback connections, showing that introducing slight structure to the network connections enables targeted attentional enhancement and the spread of feature-based attention. Together, these findings expand our understanding of how top-down attentional signals from prefrontal areas modulate sensory processing, emphasizing the interaction between high-dimensional representations in the prefrontal cortex and structured representations in early sensory areas to produce precise neural modulations required to achieve behavioral goals.
Main Content
Enter the password to open this PDF file:
-
-
-
-
-
-
-
-
-
-
-
-
-
-