A model of autonomous interactions between hippocampus and neocortex driving sleep-dependent memory consolidation

Proceedings of the National Academy of Sciences, Год журнала: 2022, Номер 119(44)

Опубликована: Окт. 24, 2022

How do we build up our knowledge of the world over time? Many theories memory formation and consolidation have posited that hippocampus stores new information, then “teaches” this information to neocortex time, especially during sleep. But it is unclear, mechanistically, how actually works—How are these systems able interact periods with virtually no environmental input accomplish useful learning shifts in representation? We provide a framework for thinking about question, neural network model simulations serving as demonstrations. The composed neocortical areas, which replay memories one another completely autonomously simulated Oscillations leveraged support error-driven leads changes representation behavior. has non–rapid eye movement (NREM) sleep stage, where dynamics between tightly coupled, helping reinstate high-fidelity versions attractors, REM more freely explore existing attractors. find alternating NREM stages, alternately focuses model’s on recent remote facilitates graceful continual learning. thus an account can without any external drive cortical protect old integrated.

Язык: Английский

---

Single and multiple systems in categorization and category learning

Nature Reviews Psychology, Год журнала: 2024, Номер 3(8), С. 536 - 551

Опубликована: Июль 22, 2024

Язык: Английский

---

Adult neurogenesis acts as a neural regularizer

Proceedings of the National Academy of Sciences, Год журнала: 2022, Номер 119(45)

Опубликована: Ноя. 2, 2022

New neurons are continuously generated in the subgranular zone of dentate gyrus throughout adulthood. These new gradually integrate into hippocampal circuits, forming naive synapses. Viewed from this perspective, these may represent a significant source “wiring” noise networks. In machine learning, such injection is commonly used as regularization technique. Regularization techniques help prevent overfitting training data and allow models to generalize learning new, unseen data. Using computational modeling approach, here we ask whether neurogenesis-like process similarly acts regularizer, facilitating generalization category task. convolutional neural network (CNN) trained on CIFAR-10 object recognition dataset, modeled neurogenesis replacement/turnover mechanism, where weights for randomly chosen small subset hidden layer were reinitialized values model learned categorize 10 different classes objects. We found that enhanced test compared networks with no neurogenesis. Moreover, neurogenic either outperformed or performed conventional (i.e., dropout, weight decay, noise). results suggest can enhance through injection, expanding roles have cognition.

Язык: Английский

---

A multilevel account of hippocampal function in spatial and concept learning: Bridging models of behavior and neural assemblies

Science Advances, Год журнала: 2023, Номер 9(29)

Опубликована: Июль 21, 2023

A complete neuroscience requires multilevel theories that address phenomena ranging from higher-level cognitive behaviors to activities within a cell. We propose an extension the level of mechanism approach where computational model cognition sits in between behavior and brain: It explains can be decomposed into lower-level component mechanisms provide richer understanding system than any alone. Toward this end, we neuron-like units using neural flocking parallels recurrent hippocampal activity. Neural coordinates collectively form mental constructs. The suggested how brain-scale populations coordinate assemblies encoding concept spatial representations why so many neurons are needed for robust performance at level. This explanation provides way understand symbol-like supported by coordinated (assemblies) formed through learning.

Язык: Английский

---

Pattern Separation and Pattern Completion Within the Hippocampal Circuit During Naturalistic Stimuli

Human Brain Mapping, Год журнала: 2025, Номер 46(2)

Опубликована: Янв. 29, 2025

ABSTRACT Pattern separation and pattern completion in the hippocampus play a critical role episodic learning memory. However, there is limited empirical evidence supporting of hippocampal circuit these processes during complex continuous experiences. In this study, we analyzed high‐resolution fMRI data from “ Forrest Gump ” open‐access dataset (16 participants) using sliding‐window temporal autocorrelation approach to investigate whether canonical (DG‐CA3‐CA1‐SUB) shows consistent with occurrence or naturalistic audio movie task. Our results revealed that when processing stimuli, DG‐CA3 pair exhibited process, whereas both CA3‐CA1 CA1‐SUB pairs showed completion. Moreover, latter half movie, observed reduction an increase pair. Overall, findings improve our understanding related natural

Язык: Английский

---

Coordinating multiple mental faculties during learning

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Фев. 13, 2025

Abstract Complex behavior is supported by the coordination of multiple brain regions. How do regions coordinate absent a homunculus? We propose achieved controller-peripheral architecture in which peripherals (e.g., ventral visual stream) aim to supply needed inputs their controllers hippocampus and prefrontal cortex) while expending minimal resources. developed formal model within this framework address how support rapid learning from few example images. The captured higher-level activity controller shaped lower-level representations, affecting precision sparsity manner that paralleled measures. In particular, peripheral encoded information extent smooth operation controller. Alternative models optimized gradient descent irrespective architectural constraints could not account for human or responses, and, typical standard deep approaches, were unstable trial-by-trial learners. While previous work offered accounts specific faculties, such as perception, attention, learning, approach step toward addressing next generation questions concerning faculties coordinate.

Язык: Английский

---

On the origin of memory neurons in the human hippocampus

Trends in Cognitive Sciences, Год журнала: 2025, Номер unknown

Опубликована: Март 1, 2025

Язык: Английский

---

Evidence for shallow cognitive maps in Schizophrenia

Cognitive Affective & Behavioral Neuroscience, Год журнала: 2025, Номер unknown

Опубликована: Март 20, 2025

Язык: Английский

---

Acute Stress Effects on Statistical Learning and Episodic Memory

Journal of Cognitive Neuroscience, Год журнала: 2024, Номер 36(8), С. 1741 - 1759

Опубликована: Янв. 1, 2024

Stress is widely considered to negatively impact hippocampal function, thus impairing episodic memory. However, the hippocampus not merely seat of Rather, it also (via distinct circuitry) supports statistical learning. On basis rodent work suggesting that stress may impair pathway involved in memory while sparing or enhancing learning, we developed a behavioral experiment investigate effects acute on both and learning humans. Participants were randomly assigned one three conditions: (socially evaluated cold pressor) immediately before ∼15 min no stress. In task, participants viewed series trial-unique scenes (allowing for encoding each image) which certain scene categories reliably followed another associations between paired categories). Memory was assessed 24 hr later isolate encoding/learning rather than retrieval. We found modest support our hypothesis can amplify learning: Only stressed advance exhibited reliable evidence across multiple measures. Furthermore, stress-induced cortisol levels predicted retention later. contrast, did differ by condition, although find preliminary promoted statistically predictable information attenuated competition encoding. Together, these findings provide initial insights into how differentially modulate processes within hippocampus.

Язык: Английский

---

Acute stress effects on statistical learning and episodic memory

Опубликована: Янв. 13, 2024

Stress is widely considered to negatively impact hippocampal function, thus impairing episodic memory. However, the hippocampus not merely seat of Rather, it also (via distinct circuitry) supports statistical learning. Based on rodent work suggesting that stress may impair pathway involved in memory while sparing or enhancing learning, we developed a behavioral experiment investigate effects acute both and learning humans. Participants were randomly assigned one three conditions: (socially evaluated cold pressor) immediately before ~15 minutes no stress. In task, participants viewed series trial-unique scenes (allowing for encoding each image) which certain scene categories reliably followed another associations between paired categories). Memory was assessed 24h later isolate encoding/learning rather than retrieval. We found modest support our hypothesis can amplify learning: only stressed advance exhibited reliable evidence across multiple measures. Further, stress-induced cortisol levels predicted retention later. contrast, did differ by condition, though find promoted statistically predictable information attenuated competition encoding. Together, these findings provide initial insights into how differentially modulate processes within hippocampus.

Язык: Английский

---