Modeling post-training memory transfer in cerebellar motor learning

  • Authors:
  • Tadashi Yamazaki;Soichi Nagao

  • Affiliations:
  • Graduate School of Informatics and Engineering, The University of Electro-Communications, Japan,Laboratory for Motor Learning Control, RIKEN Brain Science Institute, India;Laboratory for Motor Learning Control, RIKEN Brain Science Institute, India

  • Venue:
  • ICONIP'12 Proceedings of the 19th international conference on Neural Information Processing - Volume Part III
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

The cerebellum has two distinct memory sites. A single session of behavioral training forms short-term memory in the cerebellar cortex, and by repeating the training, long-term memory is formed in the cerebellar or vestibular nuclei, as if the memory is transferred from the cortex to the nuclei. We propose a simple network model of the cerebellum for the formation and transfer of motor memory. We assume a Hebbian rule with a postsynaptic gating mechanism for synaptic plasticity in the nuclei. We carry out computer simulation of gain adaptation of vestibulo-ocular reflex (VOR) and demonstrate robust memory transfer: the synaptic weight in the nuclei does not diverge to infinity. We suggest that memory transfer occurs mainly after training, not during training, and that spontaneous activity of Purkinje cells after training is necessary for memory transfer.