This is the repository for our preprint titled Reservoir-computing based associative memory and itinerancy for complex dynamical attractors. We use two different approaches - index-based reservoir computer (RC) and index-free RC - to have multifunctional recurrent neural networks with complex dynamical long-term memory states.
The training data (time series of target memory states) and trained weights of reservoir computers (RC) are shared in this OSF repository) with DOI 10.17605/OSF.IO/YXM2V. You may generate the data yourself using the codes we shared here in this GitHub repository).
They are shared under the Data folder.
They are shared under the Code-and-Results folder.
If you have any questions or would like to discuss this work further, please do not hesitate to contact me!