Takeshi Aihara

This paper investigates how CA3 recurrent connectivity affects autoassociation in a DG->CA3 model, finding that sparsity stabilizes high-connectivity networks and that binary and spiking CA3 types have opposing failure modes.

This paper uses computational models to show how pathway-specific synaptic dynamics and lateral inhibition shape frequency-dependent input integration and noise-robust pattern separation in the dentate gyrus.

The paper compares the dynamics of mutual inhibition between two real neurons with human perceptual competition, finding similar bi-stable behavior under noise and activation changes.