You must login before you can run this tool.
Recurrent networks can used as pattern auto-associators and can act as content addressable memories, and can serve as models for certain types of neurobiological memory. They are often used to model the hippocampus, a brain region closely linked to memory formation and recall. Recurrent neural networks are dynamic systems that process signals in time. The ones we consider in the tool consist of one layer of neural units that are all interconnected and are intended as stored memory states. This tool is then able to recall patterns given hints in the form of incomplete patterns. However, this tool implements asynchronous updates rather than synchronous. This makes it more stable.
Cite this work
Researchers should cite this work as follows: