We propose a computational theory of hippocampalsystem function in mediating stimulus representation in associative learning. A connectionist model based on this theory is described here, in which the hippocampal system develc^s new and adaptive stimulus repesentations which are predictive, distributed, and compressed: other cortical and cerebellar modules are presumed to use these hippocampal representations to recode their o w n stimulus representations. This computational theory can been seen as an extension and/or refinement of several prior characterizations of hippocampal function. including theories of chunking, stimulus selection, cue-configuration, and contextual coding. The theory does not address temporal aspects of hippocampal function. SimulaticMis of the intact and lesioned model provide an account of data on diverse effects of hippocampal-region lesions, including simple discrimination learning, sensory preconditioning. reversal training, latent inhibition, contextual shifts, and ccMifigural learning. Potential implications of this theory for understanding human declarative m e m o y , temporal processing. and neural mechanisms are briefly discussed.

Contemporary evidence suggests that sleep contributes to theextraction of gist from previously encoded experiences, aprocess that relies on compressed memory replay. While thefunctional significance of the time compression is not fullyunderstood, a recent ‘temporal scaffolding’ model suggestedthat compression allows associating encoded events thathappened in disparate times, a critical feature when extractinggist of a temporal nature. We examined this hypothesis usinga novel behavioral paradigm. Subjects were first presentedwith word pairs that could form a new composite word ifcombined (e.g., car, pet --> carpet), and then tested onwhether they falsely recognize seeing the composite word.When subjects napped in between exposure and testing, falsememories of composite words increased, with reaction timesfor false recognition correlating to time spent in slow wavesleep. These results confirm the functional role of timecompression in memory replay, supporting the temporalscaffolding model.

Generalization of a pattern categorization task was investigated in a simple, deterministic,inductive learning task. Each of eight patterns in a training set was specified in terms of four binary features. After subjects learned to categorize these patterns in a supervised learning paradigm they were asked generalize their knowledge by categorizing novel patterns. We analyzed both the details of the learning process as well as subjects' generalizations to novel patterns.Certain patterns in the training set were consistency found to be more difficult to learn than others.The subsequent generalizations made by subjects indicate that in spite of important individual differences, subjects showed systematic similarities in how they generalized to novel situations.The generalization performance of subjects was compared to those that could possibly be generated by a two-layer adaptive network. A comparison of network and human generalization syndicate that using a minimal network architecture is not a sufficient constraint to guarantee that a network will generalize the way humans do.

Recent work has conceptualized the brain as a network comprised of groups of sub-networks or modules. "Flexibility" of brain network(s) indexes the dynamic reconfiguration of comprising modules. Using novel techniques from dynamic network neuroscience applied to high-resolution resting-state functional magnetic resonance imaging (fMRI), the present study investigated the effects of an aerobic exercise intervention on the dynamic rearrangement of modular community structure-a measure of neural flexibility-within the medial temporal lobe (MTL) network. The MTL is one of the earliest brain regions impacted by Alzheimer's disease. It is also a major site of neuroplasticity that is sensitive to the effects of exercise. In a two-group non-randomized, repeated measures and matched control design with 34 healthy older adults, we observed an exercise-related increase in flexibility within the MTL network. Furthermore, MTL network flexibility mediated the beneficial effect aerobic exercise had on mnemonic flexibility, as measured by the ability to generalize past learning to novel task demands. Our results suggest that exercise exerts a rehabilitative and protective effect on MTL function, resulting in dynamically evolving networks of regions that interact in complex communication patterns. These reconfigurations may underlie exercise-induced improvements on cognitive measures of generalization, which are sensitive to subtle changes in the MTL.

BACKGROUND: Both sleep deficiencies and Alzheimers disease (AD) disproportionately affect older African Americans. Genetic susceptibility to AD further compounds risk for cognitive decline in this population. Aside from APOE ɛ4, ABCA7 rs115550680 is the strongest genetic locus associated with late-onset AD in African Americans. While sleep and ABCA7 rs115550680 independently influence late-life cognitive outcomes, we know too little about the interplay between these two factors on cognitive function. OBJECTIVE: We investigated the interaction between sleep and ABCA7 rs115550680 on hippocampal-dependent cognitive function in older African Americans. METHODS: One-hundred fourteen cognitively healthy older African Americans were genotyped for ABCA7 risk (n = 57 carriers of risk G allele; n = 57 non-carriers), responded to lifestyle questionnaires, and completed a cognitive battery. Sleep was assessed via a self-reported rating of sleep quality (poor, average, good). Covariates included age and years of education. RESULTS: Using ANCOVA, we found that carriers of the risk genotype who reported poor or average sleep quality demonstrated significantly poorer generalization of prior learning-a cognitive marker of AD-compared to their non-risk counterparts. Conversely, there was no genotype-related difference in generalization performance in individuals who reported good sleep quality. CONCLUSION: These results indicate that sleep quality may be neuroprotective against genetic risk for AD. Future studies employing more rigorous methodology should investigate the mechanistic role of sleep neurophysiology in the pathogenesis and progression of AD associated with ABCA7. There is also need for the continued development of non-invasive sleep interventions tailored to racial groups with specific AD genetic risk profiles.

Recent engineering considerations have prompted an improvement to the least mean squares (LMS) learning rule for training one-layer adaptive networks; incorporating a dynamically modifiable learning rate for each associative weight accellerates overall learning and provides a mechanism for adjusting the salience of individual cues (Sutton, 1992a,b). Prior research has established that the standard L M S rule can characterize aspects of animal learning (Rescorla & Wagner, 1972) and human category learning (Gluck & Bower, 1988a,b). W e illustrate here how this enhanced L M S rule is analogous to adding a cue-salience or attentional component to the psychological model, giving the network model a means for discriminating between relevant and irrelevant cues. W e then demonstrate the effectiveness of this enhanced L M S rule for modeling human performance in two non-stationary learning tasks for which the standard L M S network model fails to adequately account for the data (Hurwitz, 1990; Gluck, Glauthier, & Sutton, in preparation).

test a configural-cue network model of human classification and recognition learning based on Rescorla & Wagner's (1972) model of classical conditioning. The model extends the stimulus representation assumptions from our earlier one-layer network model (Gluck & Bower, 1988b) to include pair-wise conjunctions of features as unique cues. Like the exemplar context model of Medin & Schaffer(1978), the representational assumptions of the configural-cue network model embody an implicit exponential decay relationship between stimulus similarity and and psychological (Hamming) distance, a relationship which has received substantial independent empirical and theoretical support (Shepard, 1957,1987). In addition to results from animal learning, the model accounts for several aspects of complex human category learning, including the relationship between category similarity and linear separability in determining classification difficulty (Medin &Schwanenflugel, 1981), the relationship between classification and recognition memory for instances (Hayes-Roth & Hayes-Roth, 1977), and the impact of correlated attributes on classification (Medin, Altom, Edelson, & Freko, 1982).