A5070059815- Memory and Neural Mechanisms
- Neural dynamics and brain function
- Memory Processes and Influences
- Functional Brain Connectivity Studies
- Neural and Behavioral Psychology Studies
- Neuroscience and Neuropharmacology Research
- Neural Networks and Applications
- Topic Modeling
- EEG and Brain-Computer Interfaces
- Sleep and Wakefulness Research
- Face Recognition and Perception
- Domain Adaptation and Few-Shot Learning
- Cognitive Functions and Memory
- Child and Animal Learning Development
- Neurobiology of Language and Bilingualism
- Visual Attention and Saliency Detection
- Visual perception and processing mechanisms
- Deception detection and forensic psychology
- Neuroscience and Music Perception
- Advanced MRI Techniques and Applications
- Advanced Neuroimaging Techniques and Applications
- Sleep and related disorders
- Mental Health Research Topics
- Machine Learning and Algorithms
- Identity, Memory, and Therapy
Princeton University
2016-2025
Neuroscience Institute
2015-2024
Rutgers, The State University of New Jersey
2021
Oregon Health & Science University
2015
Ochin
2015
Charité - Universitätsmedizin Berlin
2012
University of Pennsylvania
2005
University of Colorado Boulder
2002-2005
Harvard University
1997-1998
Harvard University Press
1996-1997
The authors present a computational neural-network model of how the hippocampus and medial temporal lobe cortex (MTLC) contribute to recognition memory. hippocampal component contributes by recalling studied details. MTLC cannot support recall, but one can extract scalar familiarity signal from that tracks well test item matches items. simulations establish key differences in operating characteristics hippocampal-recall MTLC-familiarity signals identify several manipulations (e.g.,...
The authors present the context maintenance and retrieval (CMR) model of memory search, a generalized version temporal M. W. Howard J. Kahana (2002a), which proposes that search is driven by an internally maintained representation composed stimulus-related source-related features. In CMR model, organizational effects (the tendency for related items to cluster during recall sequence) arise as consequence associations between active elements features studied material. Semantic clustering due...
Here we describe a functional magnetic resonance imaging study of humans engaged in memory search during free recall task. Patterns cortical activity associated with the three categories pictures (faces, locations, and objects) were identified by pattern-classification algorithm. The algorithm was used to track reappearance these patterns period. given category's pattern correlates verbal recalls made from that category precedes event several seconds. This result is consistent hypothesis...
A growing literature suggests that the hippocampus is critical for rapid extraction of regularities from environment. Although this fits with known role in learning, it seems at odds idea specializes memorizing individual episodes. In particular, Complementary Learning Systems theory argues there a computational trade-off between learning specifics experiences and hold across those experiences. We asked whether possible to handle both statistical memorization exposed neural network model...
The hippocampus is involved in the learning and representation of temporal statistics, but little understood about kinds statistics it can uncover. Prior studies have tested various forms structure that be learned by tracking strength transition probabilities between adjacent items a sequence. We test whether learn higher-order using sequences no variance probability instead exhibit community structure. find indeed sensitive to this form structure, as revealed its representations, activity...
Understanding movies and stories requires maintaining a high-level situation model that abstracts away from perceptual details to describe the location, characters, actions, causal relationships of currently unfolding event. These models are built not only information present in current narrative, but also prior knowledge about schematic event scripts, which typical sequences encountered throughout lifetime. We analyzed fMRI data 44 human subjects (male female) presented with 16 three-minute...
Departing from traditional linguistic models, advances in deep learning have resulted a new type of predictive (autoregressive) language models (DLMs). Using self-supervised next-word prediction task, these generate appropriate responses given context. In the current study, nine participants listened to 30-min podcast while their brain were recorded using electrocorticography (ECoG). We provide empirical evidence that human and autoregressive DLMs share three fundamental computational...
Abstract The hippocampus replays experiences during quiet rest periods, and this replay benefits subsequent memory. A critical open question is how memories are prioritized for replay. We used functional magnetic resonance imaging (fMRI) pattern analysis to track item-level in the an awake period after participants studied 15 objects completed a memory test. Objects that were remembered less well replayed more period, suggesting prioritization process which weaker memories—memories most...
Humans are able to mentally construct an episode when listening another person's recollection, even though they themselves did not experience the events. However, it is unknown how strongly neural patterns elicited by mental construction resemble those found in brain of individual who experienced original Using fMRI and a verbal communication task, we traced associated with viewing specific scenes movie encoded, recalled, then transferred group naïve listeners. By comparing across 3...
Humans spontaneously organize a continuous experience into discrete events and use the learned structure of these to generalize memory. We introduce Structured Event Memory (SEM) model event cognition, which accounts for human abilities in segmentation, memory, generalization. SEM is derived from probabilistic generative dynamics defined over structured symbolic scenes. By embedding scene representations vector space parametrizing this space, combines advantages neural network approaches...