Storing static and cyclic patterns in an Hopfield neural network

While perceiving recurrent neural networks as brain-like information storing and retrieving machines, it is fundamental to explore at best these storing, indexing and retrieving capacities. This paper reviews an efficient Hebbian learning rule used to store both static and cyclic patterns in the dynamical attractors of an Hopfield neural network. A key improvement will be presented which consists in indexing the attractor information content by means of the external biases instead of, originally done, by the initial conditions. It will be shown how this addition first enables the storing of cycles sharing common patterns, then enhances the content-addressability of the learned patterns and finally allows to exploit the neural net in an hetero-associative way.