Formation and retrieval of cell assemblies in a biologically realistic spiking neural network model of area CA3 in the mouse hippocampus DOI Creative Commons
Jeffrey D. Kopsick, Joseph A. Kilgore, Gina C. Adam

et al.

Journal of Computational Neuroscience, Journal Year: 2024, Volume and Issue: 52(4), P. 303 - 321

Published: Sept. 17, 2024

Abstract The hippocampal formation is critical for episodic memory, with area Cornu Ammonis 3 (CA3) a necessary substrate auto-associative pattern completion. Recent theoretical and experimental evidence suggests that the retrieval of cell assemblies enable these functions. Yet, how are formed retrieved in full-scale spiking neural network (SNN) CA3 incorporates observed diversity neurons connections within this circuit not well understood. Here, we demonstrate data-driven SNN model quantitatively reflecting neuron type-specific population sizes, intrinsic electrophysiology, connectivity statistics, synaptic signaling, long-term plasticity mouse capable robust auto-association completion via assemblies. Our results show broad range assembly sizes could successfully systematically retrieve patterns from heavily incomplete or corrupted cues after limited number presentations. Furthermore, performance was respect to partial overlap through shared cells, substantially enhancing memory capacity. These novel findings provide computational specific biological properties produce an effective associative learning mammalian brain.

Language: Английский

Formation and retrieval of cell assemblies in a biologically realistic spiking neural network model of area CA3 in the mouse hippocampus DOI Creative Commons
Jeffrey D. Kopsick, Joseph A. Kilgore, Gina C. Adam

et al.

Journal of Computational Neuroscience, Journal Year: 2024, Volume and Issue: 52(4), P. 303 - 321

Published: Sept. 17, 2024

Abstract The hippocampal formation is critical for episodic memory, with area Cornu Ammonis 3 (CA3) a necessary substrate auto-associative pattern completion. Recent theoretical and experimental evidence suggests that the retrieval of cell assemblies enable these functions. Yet, how are formed retrieved in full-scale spiking neural network (SNN) CA3 incorporates observed diversity neurons connections within this circuit not well understood. Here, we demonstrate data-driven SNN model quantitatively reflecting neuron type-specific population sizes, intrinsic electrophysiology, connectivity statistics, synaptic signaling, long-term plasticity mouse capable robust auto-association completion via assemblies. Our results show broad range assembly sizes could successfully systematically retrieve patterns from heavily incomplete or corrupted cues after limited number presentations. Furthermore, performance was respect to partial overlap through shared cells, substantially enhancing memory capacity. These novel findings provide computational specific biological properties produce an effective associative learning mammalian brain.

Language: Английский

Citations

1