The hippocampus plays a critical role in spatial look-ahead, single-trial learning, memory consolidation, and imagination. Each of these learning dynamics depends on memory schemas. In this talk, I develop a simple hierarchical Bayesian approach define the contribution of hippocampal schemas to memory tasks. I show how the approach can be used to account for hippocampus dependent, schema-based single-trial learning and variable consolidation times. Next, I extend the schema learning approach to show how the approach can be used to predict novelty preferences in spontaneous exploration tasks. Finally, I discuss the utility of the approach for understanding individual differences in memory task performance and exploration.

When an individual receives a cochlear implant (CI) as a treatment for deafness, they undergo an intensive period of perceptual learning in order to learn how to hear. However, the process of perceptual learning is not fully understood, and there is much debate regarding the types of materials that an individual should train with, and how well they will generalize to novel tasks and listening situations. This study investigated the long-term perceptual learning of CI simulations in 194 normal hearing (NH) individuals in order to better understand the mechanisms of perceptual learning, discern which tasks provide the most robust generalization and transfer, and to assess the efficacy of a proposed training program for new adult CI users. The results will be discussed in terms of perceptual learning, interactions between general auditory and speech specific cognitive abilities, and their application to clinical practice.