Dynamic coding in a neural model of activity-silent working memory

Wijs, Chiel (2020) Dynamic coding in a neural model of activity-silent working memory. Bachelor's Thesis, Artificial Intelligence.

Preview	Text AI_BA_2020_CHIELWIJS.pdf Download (2MB) | Preview
	Text toestemming.pdf Restricted to Registered users only Download (102kB)

Abstract

Temporary strengthening of neural connections has been shown to be a mechanism capable of storing items in working memory without the need for persistent neural activity. While this type of activity-silent working memory does not depend on neural activity for the maintenance of a working memory item, it does for the encoding of that item. Moreover, the neural state, a snapshot of neural connectivity, rapidly changes during this encoding period, a phenomenon called dynamic coding. This study expanded on an existing neural model, that used short-term synaptic plasticity to model activity-silent working memory, by implementing a simplification of neural connectivity from the human visual system into the sensory part of the model. Analysis of the functional model shows that the resulting distributed response latency, in combination with the short-term synaptic plasticity mechanism, produces dynamic coding measures similar to those seen from analysis of human EEG data. Performance of the model decreases slightly compared to human performance when items are required to be remembered over a longer period of time. Implementation of attention control could reduce the differences we see in dynamic coding measures between the model and human subjects.

Item Type:	Thesis (Bachelor's Thesis)
Supervisor name:	Borst, J.P.
Degree programme:	Artificial Intelligence
Thesis type:	Bachelor's Thesis
Language:	English
Date Deposited:	10 Jul 2020 11:09
Last Modified:	10 Jul 2020 11:09
URI:	https://fse.studenttheses.ub.rug.nl/id/eprint/22521

Actions (login required)

View Item