Related Experiment Videos
A unified model of spatial and episodic memory.
Edmund T Rolls1, Simon M Stringer, Thomas P Trappenberg
1Department of Experimental Psychology, University of Oxford, South Parks Road, UK. edmund.rolls@psy.ox.ac.uk
Proceedings. Biological Sciences
|June 14, 2002
Summary
A unified model demonstrates how the hippocampus stores both episodic memories and spatial representations. This network is essential for remembering objects and their locations, integrating memory and spatial theories.
Area of Science:
- Neuroscience
- Cognitive Science
- Computational Neuroscience
Background:
- Medial temporal lobe structures, including the hippocampus, are crucial for episodic memory and spatial navigation.
- Previous research has investigated these functions separately.
Purpose of the Study:
- To propose a unified computational model for medial temporal lobe function.
- To demonstrate how a single network can represent both discrete episodic memories and continuous spatial information.
Main Methods:
- Development of a single recurrent attractor network model.
- Simulation of the network's capacity to store discrete and continuous representations.
Main Results:
- The recurrent attractor network successfully stores both episodic memories and spatial representations.
- Integrating object information with spatial location storage necessitates a unified network structure.
Conclusions:
- Episodic memory and spatial navigation theories of medial temporal lobe function can be reconciled within a single model.
- Recurrent attractor networks provide a viable mechanism for dual memory and spatial representation.