Abstract:It is widely believed that memory storage depends on activity-dependent synaptic modifications. Classical studies of learning and memory in neural networks describe synaptic efficacy either as continuous or discrete. However, recent results suggest an intermediate scenario in which synaptic efficacy can be described by a continuous variable, but whose distribution is peaked around a small set of discrete values. Motivated by these results, we explored a model in which each synapse is described by a continuous variable that evolves in a potential with multiple minima. External inputs to the network can switch synapses from one potential well to another. Our analytical and numerical results show that this model can interpolate between models with discrete synapses which correspond to the deep potential limit, and models in which synapses evolve in a single quadratic potential. We find that the storage capacity of the network with double well synapses exhibits a power law dependence on the network size, rather than the logarithmic dependence observed in models with single well synapses. In addition, synapses with deeper potential wells lead to more robust information storage in the presence of noise. When memories are sparsely encoded, the scaling of the capacity with network size is similar to previously studied network models in the sparse coding limit. A long-lasting question in neuroscience is whether synaptic efficacies should be described as continuous variable or discrete variables. Recent experiments indicate that it is a combination of both: synaptic efficacy changes continuously, but its distribution peaks at several discrete values. In this study, we introduce a synapse model described by a double well potential, and investigate the memory properties of networks of neurons connected with such synapses. Our results show in networks with a bimodal weight distribution, the storage capacity depends on network size as a power law. In addition, we demonstrate that networks with such synapses store information more robustly in the presence of noise, compared to networks with synapses with a single well potential.

Simplicial Hopfield networks

Universal Hopfield Networks: A General Framework for Single-Shot Associative Memory Models

Long Sequence Hopfield Memory

A Hopfield-like model with complementary encodings of memories

Long sequence Hopfield memory*

Large Associative Memory Problem in Neurobiology and Machine Learning

Storage Capacity of the Hopfield Network Associative Memory

Memorizing morph patterns in small-world neuronal network

Memory Dynamics in Attractor Networks with Saliency Weights.

Sparse and Structured Hopfield Networks

Capacity of the Hebbian-Hopfield network associative memory

Heteroclinic cycles in Hopfield networks

Modern Hopfield Networks meet Encoded Neural Representations -- Addressing Practical Considerations

Storing overlapping associative memories on latent manifolds in low-rank spiking networks

Dense Associative Memory Through the Lens of Random Features

On the use of associative memory in Hopfield networks designed to solve propositional satisfiability problems

Hopfield-Fenchel-Young Networks: A Unified Framework for Associative Memory Retrieval

Attractor neural networks with double well synapses

The capacity of the Hopfield associative memory

Associative memory realized by a reconfigurable memristive Hopfield neural network