Understanding the neural mechanisms underlying metacontrol and conflict regulation is crucial for insights into cognitive flexibility and persistence. This study employed electroencephalography (EEG), EEG-beamforming and directed connectivity analyses to explore how varying metacontrol states influence conflict regulation at a neurophysiological level. Metacontrol states were manipulated by altering the frequency of congruent and incongruent trials across experimental blocks in a modified flanker task, and both behavioral and electrophysiological measures were analyzed. Behavioral data confirmed the experimental manipulation's efficacy, showing an increase in persistence bias and a reduction in flexibility bias during increased conflict regulation. Electrophysiologically, theta band activity paralleled the behavioral data, suggesting that theta oscillations reflect the mismatch between expected metacontrol bias and actual task demands. Alpha and beta band dynamics differed across experimental blocks, though these changes did not directly mirror behavioral effects. Post-response alpha and beta activity were more pronounced in persistence-biased states, indicating a neural reset mechanism preparing for future cognitive demands. By using a novel artificial neural networks method, directed connectivity analyses revealed enhanced inter-regional communication during persistence states, suggesting stronger top-down control and sensorimotor integration. Overall, theta band activity was closely tied to metacontrol processes, while alpha and beta bands played a role in resetting the neural system for upcoming tasks. These findings provide a deeper understanding of the neural substrates involved in metacontrol and conflict monitoring, emphasizing the distinct roles of different frequency bands in these cognitive processes.