No description available.

Thesis advisor: Elizabeth A. Kensinger / A key feature of human memory is the ability to remember not only discrete pieces of information but also to form novel associations between them. A special type of association, called an "integration", can be formed when the pieces are encoded as a single representation in memory (Wollen, Weber and Lowry, 1972; Murray and Kensinger, 2012). The work presented here investigates what neural mechanisms underlie the formation and subsequent retrieval of integrated mental images in younger adulthood (individuals aged 18-30), whether those mechanisms differ based on the emotional content of the integration, and whether older and younger adults generate and remember emotional integrations differently from one another. I show that younger adults utilize two different routes to form integrations, depending on their emotional content: a rapid, perceptually-supported route that allows for fast integration of emotional pairs but that leads to poor downstream memory for the associates, and a slow, conceptually-supported route for neutral pairs that takes more time but that leads to strong downstream memory. Conversely, older adults utilize slow, controlled processing of emotional integrations that leads to strong memory, but they fail to produce durable memory for non-emotional pairs due to age-related associative deficits. Together, these results highlight differences both within and between age groups in the formation and retrieval of emotional and non-emotional integrations, and suggest a circumstance - integration of emotional pairs - in which older adults can overcome previously reported age-related deficits. / Thesis (PhD) — Boston College, 2013. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Psychology.

Young adults often outperform older adults on tests of associative memory, however, the source of this age-related associative memory deficit is still under debate. There are two main non-mutually exclusive hypotheses: 1) impaired binding processes (i.e. creating and retrieving links between units of information) and; 2) impaired strategic processes (i.e. cognitive control processes that support encoding and retrieval). Although both components are thought to contribute uniquely and interact to support associative memory, they have rarely been studied together. The primary goal of this dissertation is to further characterize associative memory deficits in healthy aging by measuring and controlling binding and strategic processes. Specifically, in this series of three experiments, we studied these two components concurrently by varying the level of demands on binding (i.e. comparing memory for different types of associations) and strategic processes (i.e. varying demands on self-initiated processes). A total N of 97 young adults and 94 older adults studied lists of object-pairs and object-location pairs under intentional encoding conditions. Demands on self-initiated processes were manipulated by increasing the number of foils at test (Experiment 1: 4 alternative forced-choice (AFC), vs. Experiment 2 & 3: 20AFC), and by providing strategy instructions in Experiment 3. We measured the production of strategies with trial-by-trial self-report. In all three experiments, we found that young adults outperformed older adults on object-object memory, but not on object-location memory. Older adults were just as proficient as young adults in generating strategies at study. This remained true even when demands on self-initiated processes increased. However, we found in all three experiments that young adults had greater strategy effectiveness (i.e. accuracy on pairs encoded with a strategy) on the object-object test. In contrast, performance on the object-location task was found to be less related to strategies. Our findings suggest that not all associations are equally affected by aging and that even when strategy production is equivalent between age groups older adults can still be impaired on associative memory. The secondary goal of this dissertation was to explore the contribution of individual variability in age, general cognitive functioning, meta-memory and executive functioning on object-object and object-location memory, strategy production, and strategy effectiveness. Our results highlight the important contribution of executive functioning over and above any effects of age in explaining age-related associative memory decline.

This thesis is concerned with one important question in artificial neural networks, that is, how biologically inspired connectivity of a network affects its associative memory performance. In recent years, research on the mammalian cerebral cortex, which has the main responsibility for the associative memory function in the brains, suggests that the connectivity of this cortical network is far from fully connected, which is commonly assumed in traditional associative memory models. It is found to be a sparse network with interesting connectivity characteristics such as the “small world network” characteristics, represented by short Mean Path Length, high Clustering Coefficient, and high Global and Local Efficiency. Most of the networks in this thesis are therefore sparsely connected. There is, however, no conclusive evidence of how these different connectivity characteristics affect the associative memory performance of a network. This thesis addresses this question using networks with different types of connectivity, which are inspired from biological evidences. The findings of this programme are unexpected and important. Results show that the performance of a non-spiking associative memory model is found to be predicted by its linear correlation with the Clustering Coefficient of the network, regardless of the detailed connectivity patterns. This is particularly important because the Clustering Coefficient is a static measure of one aspect of connectivity, whilst the associative memory performance reflects the result of a complex dynamic process. On the other hand, this research reveals that improvements in the performance of a network do not necessarily directly rely on an increase in the network’s wiring cost. Therefore it is possible to construct networks with high associative memory performance but relatively low wiring cost. Particularly, Gaussian distributed connectivity in a network is found to achieve the best performance with the lowest wiring cost, in all examined connectivity models. Our results from this programme also suggest that a modular network with an appropriate configuration of Gaussian distributed connectivity, both internal to each module and across modules, can perform nearly as well as the Gaussian distributed non-modular network. Finally, a comparison between non-spiking and spiking associative memory models suggests that in terms of associative memory performance, the implication of connectivity seems to transcend the details of the actual neural models, that is, whether they are spiking or non-spiking neurons.

Thesis advisor: Gorica D. Petrovich / Learned associations are formed when cues from the environment are paired with biologically important events and can later drive appetitive and aversive behaviors. These behaviors can persist and reappear after extinction because the original learned associations continue to exist. In particular, cues previously associated with food can later stimulate appetite and food consumption in the absence of hunger. Renewal, or reinstatement, of extinguished conditioned behaviors may help explain the mechanisms underlying persistent responding to food cues and difficulty associated with changing unhealthy eating habits. The aim of this dissertation was to determine key components in the neural circuitry mediating renewal of responding to food cues. The main focus was on the ventromedial prefrontal cortex (vmPFC; includes the infralimbic (ILA) and prelimbic (PL) areas) because that region was selectively recruited during context-dependent renewal (Chapter 3). In all of the experiments, the behavior and neural substrates of male and female rats were compared. It was important to examine both males and females because sex differences in context-mediated renewal were recently established: males consistently show renewal responding while females fail to do so (Chapters 2 and 3). The first study in this dissertation examined whether behavioral sex differences were driven by estradiol (Chapter 2) and whether the vmPFC is recruited during renewal responding (Fos induction; Chapter 3). Then, to establish the vmPFC is causal in driving the behavioral responding during renewal in a sex-specific way (Chapter 4), the vmPFC was silenced in males and stimulated it in females. This was accomplished using a chemogenetic methodology, DREADDs (Designer Receptors Exclusively Activated by Designer Drugs). Inhibiting the vmPFC in males blocks renewal responding. Reversely, stimulating the vmPFC in females resulted in renewal of responding. To determine key components of the vmPFC circuitry mediating renewal and whether these were different in males and females the experiments in Chapter 5 examined activation of PL inputs using a retrograde tract tracing combined with Fos detection design. The pathways to the PL from the ventral hippocampal formation (subiculum and CA1), the thalamus (anterior paraventricular nucleus), and the amygdala (anterior basolateral nucleus) were recruited in males and not recruited in females. This lack of recruitment could explain the lack of behavioral responding during renewal for females. Taken together, there are distinct and sex-specific circuitries recruited during context-mediated renewal. The findings from these experiments advanced our understanding of the neural mechanisms underlying sex differences in associative memory and contextual processing. They are also important for our understanding of the resilience of food cue to influence our consumption and diet choices. / Thesis (PhD) — Boston College, 2017. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Psychology.

During the past decade, a new class of knowledge representation has emerged known as structured distributed representation (SDR). A number of schemes for encoding and manipulating such representations have been developed; e. g. Pollack's Recursive Auto-Associative Memory (RAAM), Kanerva's Binary Spatter Code (BSC), Gayler's MAP encoding, and Plate's Holographically Reduced Representations (HRR). All such schemes encode structural information throughout the elements of high dimensional vectors, and are manipulated with rudimentary algebraic operations. <br /><br /> Most SDRs are very compact; components and compositions of components are all represented as fixed-width vectors. However, such compact compositions are unavoidably noisy. As a result, resolving constituent components requires a cleanup memory. In its simplest form, cleanup is performed with a list of vectors that are sequentially compared using a similarity metric. The closest match is deemed the cleaned codevector. <br /><br /> While SDR schemes were originally designed to perform cognitive tasks, none of them have been demonstrated in a neurobiologically plausible substrate. Potentially, mathematically proven properties of these systems may not be neurally realistic. Using Eliasmith and Anderson's (2003) Neural Engineering Framework, I construct various spiking neural networks to simulate a general cleanup memory that is suitable for many schemes. <br /><br /> Importantly, previous work has not taken advantage of parallelization or the high-dimensional properties of neural networks. Nor have they considered the effect of noise within these systems. As well, additional improvements to the cleanup operation may be possible by more efficiently structuring the memory itself. In this thesis I address these lacuna, provide an analysis of systems accuracy, capacity, scalability, and robustness to noise, and explore ways to improve the search efficiency.

During the past decade, a new class of knowledge representation has emerged known as structured distributed representation (SDR). A number of schemes for encoding and manipulating such representations have been developed; e. g. Pollack's Recursive Auto-Associative Memory (RAAM), Kanerva's Binary Spatter Code (BSC), Gayler's MAP encoding, and Plate's Holographically Reduced Representations (HRR). All such schemes encode structural information throughout the elements of high dimensional vectors, and are manipulated with rudimentary algebraic operations. <br /><br /> Most SDRs are very compact; components and compositions of components are all represented as fixed-width vectors. However, such compact compositions are unavoidably noisy. As a result, resolving constituent components requires a cleanup memory. In its simplest form, cleanup is performed with a list of vectors that are sequentially compared using a similarity metric. The closest match is deemed the cleaned codevector. <br /><br /> While SDR schemes were originally designed to perform cognitive tasks, none of them have been demonstrated in a neurobiologically plausible substrate. Potentially, mathematically proven properties of these systems may not be neurally realistic. Using Eliasmith and Anderson's (2003) Neural Engineering Framework, I construct various spiking neural networks to simulate a general cleanup memory that is suitable for many schemes. <br /><br /> Importantly, previous work has not taken advantage of parallelization or the high-dimensional properties of neural networks. Nor have they considered the effect of noise within these systems. As well, additional improvements to the cleanup operation may be possible by more efficiently structuring the memory itself. In this thesis I address these lacuna, provide an analysis of systems accuracy, capacity, scalability, and robustness to noise, and explore ways to improve the search efficiency.

No description available.

It is well established that older adults show a deficit in episodic memory. The associative deficit hypothesis (ADH) (Naveh-Benjamin, 2000) suggests that an age-related reduced ability to create links between units of information is a major contributor to the episodic deficit. It has been a robust finding that older adults show a disproportionate decline in associative memory relative to item memory when compared to young adults. Previous researchers have investigated the role of the frontal lobes (FL) by studying the effect of reduced attentional resources in the associative deficit. However, they have not found that divided attention in young adults produces the disproportionate associative decline seen in aging and it is thought that some cognitive process other than the allocation of attentional resources may contribute to the associative deficit. The present study intended to use a divided attention (DA) task that also engages medial temporal brain regions (MTL) in order to tax additional parts of the network involved in creating associations and provide indirect support for the role of the MTL in the associative deficit. However, the associative memory deficit in older adults was not replicated due to unique poor associative memory performance of some young adults in the study. Analyses excluding these participants show support for the role of the MTL in the associative deficit. However, the young poor performers may provide support for the role of FL function in the associative deficit and show that poor associative memory may not be limited to the older adult cohort.

This thesis addresses the problem of how the dendritic structure and other morphological properties of the neuron can determine its pattern recognition performance. The techniques used in this work for generating dendritic trees with different morphologies included the following three methods. Firstly, dendritic trees were produced by exhaustively generating every possible morphology. Where this was not possible due to the size of morphological space, I sampled systematically from the possible morphologies. Lastly, dendritic trees were evolved using an evolutionary algorithm, which varied existing morphologies using selection, mutation and crossover. From these trees, I constructed full compartmental conductance-based models of neurons. I then assessed the performance of the resulting neuronal models by quantifying their ability to discriminate between learned and novel input patterns. The morphologies generated were tested in the presence and absence of active conductances. The results have shown that the morphology does have a considerable effect on pattern recognition performance. In fact, neurons with a small mean depth of their dendritic tree are the best pattern recognizers. Moreover, the performance of neurons is anti-correlated with mean depth. Interestingly, the symmetry of the neuronal morphology does not correlate with performance. This research has also revealed that the evolutionary algorithm could find effective morphologies for both passive models and models with active conductances. In the active model, there was a considerable change in the performance of the original population of neurons, which largely resulted from changes in the morphological parameters such as dendritic compartmental length and tapering. However, no single parameter setting guaranteed good neuronal performance; in three separate runs of the evolutionary algorithm, different sets of well performing parameters were found. In fact, the evolved neurons performed at least five times better than the original hand-tuned neurons. In summary, the combination of morphological parameters plays a key role in determining the performance of neurons in the pattern recognition task and the right combination produces very well performing neurons.