Tatsuya Haga

This paper introduces a momentum Hopfield model to explain how hippocampal attractor dynamics enable prioritized memory sampling, enhancing reinforcement learning for spatial navigation.

The paper shows that in mice, preconfigured engram cell ensembles during sleep correlate with learning, and postlearning synaptic changes prepare for new memories.

This paper introduces a neural representation model that unifies spatial and semantic computations by linking spatial navigation value functions to word embedding information measures.

This study presents a cortical model of local-global inhibition balance to investigate how inhibitory neuron diversity affects memory retrieval in associative memory structures.

This paper introduces a model called Laplacian associative memory to simulate graph-based hierarchical computation in the brain, showing it generates multiscale representations similar to graph Laplacian eigenvectors and exhibits hippocampal-like chunking.

The paper shows that individual differences in decision making under uncertainty are linked to variability in neural dynamics in the medial frontal cortex of rats.

The paper presents a computational model of dendritic processing in CA3 neurons that uses preplay sequences for one-shot learning, showing how dendritic computations enable stable memory formation.

The paper shows that symmetric STDP with short-term plasticity biases synaptic transmission during reverse replay, enhancing goal-directed spatial memory in a T-maze.

This paper introduces a method to detect cell-assembly sequences in neural activity without external references, revealing new structures in rat prefrontal cortex during goal-directed behavior.