Novel verbal instructions recruit abstract neural patterns of time-variable information dimensionality

Abstract

AbstractHuman performance is endowed by neural representations of information that is relevant for the task, some of which are also activated in a preparatory fashion to optimize later execution. Most studies to date have focused on highly practiced actions, leaving largely unaddressed the configuration of neural information in novel settings, where unique task sets have to be generated from scratch. Using electroencephalography (EEG), this study investigated the dynamics of the content and geometry reflected on the neural patterns of control representations during novel instructed behavior. We designed a verbal instruction paradigm where each trial involved novel combinations of multi-component task information. By manipulating three task-relevant factors, we observed multiplexed coding of information throughout the trial, during both preparation and implementation stages. The temporal profiles were consistent with a hierarchical structure: higher-level task information was coded in a sustained manner, while lower-level variables were so more transiently. Data showed both high dimensionality and abstraction, particularly during instruction encoding and target processing. Our results suggest that whenever task content could be recovered from neural patterns of activity, it was structured in an abstract format, with an underlying structure that favored generalization. During target processing, where potential interference across factors increased, orthogonal configurations also appeared. Overall, our findings uncover the dynamic manner with which control representations operate during novel scenarios, with changes in dimensionality and abstraction adjusting to task needs.