Our sensory systems are responsible for maintaining an accurate representation of the physical world around us while also quickly alerting us to changes in the environment. One theory of how the sensory systems perform such computations is the predictive coding theory, which posits that sensory systems maintain a stable internal representation of the physical world and propagate error signals when mismatches between incoming stimuli occur with said internal representation. The biological mechanisms underlying and supporting this theory remain open questions. In Chapter 2, we tested whether neurons in auditory cortex (AC) are sensitive to higher-order stimulus statistics using custom chord stimuli presented in temporally regular and random sequences while recording firing rates electrophysiologically. We find that neuronal firing rates in response to context transitions can be explained by a two-step model in which adaptation to frequency is separated from adaptation to regularity. In Chapter 3, we use optogenetics and 2-photon imaging to investigate which specific interneuron subtypes contribute to novelty responses in regular and random contexts. By optogenetically inactivating parvalbumin- (PV), somatostatin- (SST), and vasoactive intestinal protein- (VIP) expressing interneurons, we find that VIPs selectively enhance the novelty response in regular but not random contexts. SSTs and PVs, on the other hand, broadly suppress the novelty response independent of the temporal context. These experiments show that neurons in AC adapt to higher-order stimulus statistics and the mechanisms supporting such responses.