Associative Memory System via Threshold Linear Networks
This work addresses memory retrieval for AI systems, but it appears incremental as it builds on existing auto-associative memory models with added guarantees.
The authors tackled the problem of auto-associative memory in stochastic environments by proposing an online system with sequential memory formation, achieving successful pattern reconstruction from corrupted inputs in simulations.
Humans learn and form memories in stochastic environments. Auto-associative memory systems model these processes by storing patterns and later recovering them from corrupted versions. Here, memories are learned by associating each pattern with an attractor in a latent space. After learning, when (possibly corrupted) patterns are presented to the system, latent dynamics facilitate retrieval of the appropriate uncorrupted pattern. In this work, we propose a novel online auto-associative memory system. In contrast to existing works, our system supports sequential memory formation and provides formal guarantees of robust memory retrieval via region-of-attraction analysis. We use a threshold-linear network as latent space dynamics in combination with an encoder, decoder, and controller. We show in simulation that the memory system successfully reconstructs patterns from corrupted inputs.