Associative Long Short-Term Memory
This addresses a bottleneck in memory-augmented neural networks for AI applications, though it appears incremental as it builds on existing Holographic Reduced Representations and LSTM networks.
The paper tackles the problem of limited capacity and noisy retrieval in associative memory for recurrent neural networks by introducing a method that creates redundant copies of stored information, resulting in faster learning on memorization tasks.
We investigate a new method to augment recurrent neural networks with extra memory without increasing the number of network parameters. The system has an associative memory based on complex-valued vectors and is closely related to Holographic Reduced Representations and Long Short-Term Memory networks. Holographic Reduced Representations have limited capacity: as they store more information, each retrieval becomes noisier due to interference. Our system in contrast creates redundant copies of stored information, which enables retrieval with reduced noise. Experiments demonstrate faster learning on multiple memorization tasks.