The brain receives numerous sensory stimuli in any given moment. Detecting the structure in these incoming inputs helps understand which stimuli are relevant and which not, as well as make predictions about future stimuli. Detecting the structure in the incoming stimuli, a process known as statistical learning, requires detecting and learning the associations between stimuli, associations that occur across varied time intervals, and often in the absence of a reinforcer. This process is known as statistical learning and is particularly important in the auditory domain. Using a combination of behaviour and electrophysiology, we study the role of auditory subcortico-cortical loops in statistical learning and the effect of this learning on sensory gating.