| dc.description.abstract | Behavioral flexibility requires the selective processing of stimulus inputs as it relates to current goals. Such selective processing is instantiated in dispersed brain circuits, including the anterior cingulate cortex, lateral prefrontal cortex, and striatum. Many studies have delineated specific functional roles for these areas in the process of learning stimulus values, computing deviations from expectations, and adjusting to changing task demands. Such values are conveyed via an as as-of-yet underspecified neural code. One possibility is that band-limited rhythmic activity can support and complement encoding of learning variables, either through increases in the signal to noise ratio, or by providing an independent channel for encoding information. Evidence for such a possibility comes from the fact that oscillatory activity can support encoding schemes that allow for selective, dynamic routing of information across fronto-striatal circuits. Such oscillations are supported by specific cellular motifs that are localized to specific cortical lamina. However, even within single cells, there exists a wide range of dynamic states, including but not limited to Poisson firing, or non-Poisson burst firing. It is unknown how these neuronal activity states are related to network-level oscillatory activity during information encoding and transmission. This thesis addresses these questions by looking at how oscillatory activity is related to different neuronal firing dynamics, and if phase-of-firing activity encodes information related to learning. We further develop a novel experimental suite to help extend these insights beyond reduced-complexity stimuli and into the realm of more naturalistic tasks. | |
