Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network

  Advanced Search  

Limits on the memory storage capacity of bounded synapses

Summary: Limits on the memory storage capacity of
bounded synapses
Stefano Fusi & L F Abbott
Memories maintained in patterns of synaptic connectivity are rapidly overwritten and destroyed by ongoing plasticity related to the
storage of new memories. Short memory lifetimes arise from the bounds that must be imposed on synaptic efficacy in any realistic
model. We explored whether memory performance can be improved by allowing synapses to traverse a large number of states
before reaching their bounds, or by changing the way these bounds are imposed. In the case of hard bounds, memory lifetimes
grow proportional to the square of the number of synaptic states, but only if potentiation and depression are precisely balanced.
Improved performance can be obtained without fine tuning by imposing soft bounds, but this improvement is only linear with
respect to the number of synaptic states. We explored several other possibilities and conclude that improving memory
performance requires a more radical modification of the standard model of memory storage.
The idea that memories are stored through long-lasting modifications
of synaptic strengths within neural circuits has become a basic postulate
of neuroscience. Experimentalists have made great strides in providing
evidence for this viewpoint1,2, and theorists have added further sup-
port3,4. In particular, theorists have shown that models using long-term
synaptic plasticity as a storage mechanism can retain enormous
numbers of memories for long periods of time. The models used to
establish this result involve many simplifying assumptions and not-so-
realistic features, but it has generally been believed that their impressive


Source: Andrzejak, Ralph Gregor - Departament de Tecnologia, Universitat Pompeu Fabra
Fusi, Stefano - Institute für Neuroinformatik, Universität Zürich & Center for Theoretical Neuroscience, Columbia University


Collections: Biology and Medicine; Computer Technologies and Information Sciences