Abstract
Statistical learning, sensory-driven unsupervised learning of repeating patterns, must coexist with ongoing homeostatic plasticity that is responsible for the necessary balance of activity in the brain; however, the mechanisms that facilitate these interactions are not clear. While models of both statistical learning, a form of associative plasticity, and homeostatic plasticity have primarily focused on excitatory cells and their synaptic changes, inhibition may play a key role in facilitating the balance between homeostatic plasticity and statistical learning. Here, we review the inhibitory synaptic, cellular, and network mechanisms underlying homeostatic and associative plasticity in rodents and propose a model in which localized inhibition, provided by diverse interneuron types, supports both statistical learning and homeostatic plasticity, as well as the interactions between them.