1 |
Cognitive Science Honors the Memory of Jeffrey Elman
|
|
|
|
In: MIT Press (2021)
|
|
BASE
|
|
Show details
|
|
2 |
Synesthetes perseverate in implicit learning: Evidence from a non-stationary statistical learning task
|
|
|
|
BASE
|
|
Show details
|
|
5 |
The effect of Zipfian frequency variations on category formation in adult artificial language learning
|
|
|
|
BASE
|
|
Show details
|
|
6 |
Indexical and linguistic processing by 12-month-olds : discrimination of speaker, accent and vowel differences
|
|
|
|
BASE
|
|
Show details
|
|
7 |
Learning and processing of perceptual confusability and the mapping of form to meaning
|
|
|
|
BASE
|
|
Show details
|
|
8 |
Sampling over Nonuniform Distributions: A Neural Efficiency Account of the Primacy Effect in Statistical Learning
|
|
|
|
BASE
|
|
Show details
|
|
9 |
Statistical learning: A powerful mechanism that operates by mere exposure
|
|
|
|
BASE
|
|
Show details
|
|
10 |
Is synesthesia more than unusual associations? : examining cue combination and various forms of learning in synesthetes.
|
|
|
|
BASE
|
|
Show details
|
|
11 |
Indexical and linguistic processing in infancy : discrimination of speaker, accent and vowel differences
|
|
|
|
BASE
|
|
Show details
|
|
12 |
Learning across space, time, and input modality : towards an integrative, domain-general account of the neural substrates underlying visual and auditory statistical learning
|
|
|
|
Abstract:
Thesis (Ph. D.)--University of Rochester. Department of Brain and Cognitive Sciences, 2014. ; In the present work, we detail a set of experiments aimed at elucidating the neural mechanisms underpinning statistical learning across space, time, and input modality. Specifically, we have employed functional magnetic resonance imaging (fMRI) to test directly the hypothesis that distributional learning recruits a common set of neural substrates, regardless of the domain of the structure to be acquired. In experiment 1, we made use of an intermittent testing design in order to monitor changes in learning over time during a word segmentation task. By relating fluctuations in behavioral performance with differences in the magnitude of neural response across exposure runs, we demonstrated the involvement of a fronto-subcortical network of regions supporting statistical learning, with peak activation in the left inferior frontal gyrus. In experiment 2, we investigated the brain basis of learning when we shifted not only the modality of the input, but also its spatiotemporal properties. In contrast to the sequentially-ordered segmentation task in the previous study, experiment 2 sought to uncover the regions recruited during the acquisition of simultaneously-presented visuospatial patterns. Again capitalizing on inter and intra-subject variability in behavior, we revealed involvement of a parallel fronto-subcortical circuit that additionally encompassed bilateral amygdala. A further connectivity analysis using seeds within this network made clear a striking pattern: for each univariate activation peak, functional coupling was stronger in the first exposure run relative to the last exposure run. Finally, experiment 3 combined sequential learning in the auditory and visual modalities. We exposed participants to one of two carefully matched conditions. In the auditory condition, they completed a word segmentation task similar to the one described in experiment 1. In the visual condition, participants were exposed to an identical language, but one in which each syllable was replaced with a shape. Intermittent test scores showed behavioral performance that was slower to reach above-chance levels and less robust than in the segmentation task of the first study. Neuroimaging analyses revealed hippocampal, not fronto-subcortical, involvement correlated with changes in performance, and we discussed this finding in light of crucial differences between the rates of learning in experiments 1 and 3. Similar to the results of experiment 2, a subsequent functional connectivity analysis suggested greater interregional coherence in the earliest phases of learning. Linking together results from the three experiments, we propose a two-part mechanism to the neural basis of statistical learning. We posit that the brain, when confronted with structured stimuli, immediately engages widespread network of frontal, subcortical, and hippocampal regions. Over time, this network narrows, and the substrates best suited to perform the computations required of task at hand assume the processing burden. With influence from well-known proposals of the computational architecture underlying learning in the brain (e.g., Atallah, Frank, & O’Reilly, 2004; McClelland, McNaughton, & O’Reilly, 1995), we suggest that prefrontal cortex and basal ganglia form a complementary circuit best suited for the maintenance and updating of internal representations, while medial temporal regions are best suited for calculating the rapid element-to-element associations crucial to the earliest stages of a statistical learning task.
|
|
Keyword:
Basal ganglia; fMRI; functional connectivity; hippocampus; IFG; Statistical learning
|
|
URL: http://hdl.handle.net/1802/28854
|
|
BASE
|
|
Hide details
|
|
13 |
Infants' goal anticipation during failed and successful reaching actions
|
|
|
|
BASE
|
|
Show details
|
|
14 |
Phonetic Category Learning and Its Influence on Speech Production
|
|
|
|
BASE
|
|
Show details
|
|
15 |
Rational perspectives on the role of stimulus order in human cognition
|
|
|
|
BASE
|
|
Show details
|
|
19 |
The neural correlates of statistical learning in a word segmentation task: An fMRI study
|
|
|
|
In: ISSN: 0093-934X ; Brain and Language, Vol. 127, No 1 (2013) pp. 46-54 (2013)
|
|
BASE
|
|
Show details
|
|
20 |
Infants’ goal anticipation during failed and successful reaching actions
|
|
|
|
BASE
|
|
Show details
|
|
|
|