Transfer between long-term and short-term memory using Conceptors
This work addresses memory modeling in neural networks, but it appears incremental as it builds on existing conceptor methods without demonstrating broad SOTA impact.
The authors tackled the problem of modeling working memory by combining short-term and long-term components in a recurrent neural network, using conceptors to store long-term patterns and a gated reservoir for short-term storage, enabling bidirectional transfer and combination of memories.
We introduce a recurrent neural network model of working memory combining short-term and long-term components. e short-term component is modelled using a gated reservoir model that is trained to hold a value from an input stream when a gate signal is on. e long-term component is modelled using conceptors in order to store inner temporal patterns (that corresponds to values). We combine these two components to obtain a model where information can go from long-term memory to short-term memory and vice-versa and we show how standard operations on conceptors allow to combine long-term memories and describe their effect on short-term memory.