Anders Lansner: Theories, models and simulations of human working memory
Time: Fri 2020-11-06 11.00 - 12.00
Location: Zoom meeting ID: 636 5838 1373
Participating: Anders Lansner, KTH & SU
Abstract
I will first talk generally about human memory systems and working memory (WM) in particular. Then the ongoing paradigm shift in our understanding of WM and the two competing theories, i.e. persistent activity vs synaptic WM, will be explained. The Hopfield network style neural network models of cortical memory systems will be described, including versions with non-spiking and spiking neural units. Two simulation examples will be given based on synaptic WM, more precisely Hebbian fast synaptic plasticity. The first is a simulation of human wordlist learning which demonstrates and suggests explanations for primacy and recency in free recall. The other is a model of the interaction between WM and LTM (long-term memory) based on an “indexing theory” of working memory featuring two memory areas in prefrontal and parieto-temporal cortices. Finally, I will speculate about the relation between working memory processes and human thinking processes.
Anders Lansner,
Professor in Computer Science at Stockholm University Mathematics and KTH Royal Institute of Technology, Department of Computational Science and Technology (EECS/CST) Computational Brain Science Lab