workmate-match | Deep WorkMATe model with learned Biologically

workmate-match is a Python library. workmate-match has no bugs, it has no vulnerabilities and it has low support. However workmate-match build file is not available. You can download it from GitHub.

The project entails the extension of a Biologically Plausible Reinforcement Learning Agent (WorkMATe[1]) to Deep Learning. This enables the learning of memory representations of natural scenes (CIFAR-10 dataset) instead of short binary vectors. Secondly, research was done into neural mechanisms underlying matching between sensory and memory inputs. Results from an earlier neural comparator [2] were replicated and adapted to learn within the RL agent. It was found that two sequential phases of learning are needed:. [1]"Flexible Working Memory through Selective Gating and Attentional Tagging" Wouter Kruijne, Sander M Bohte, Pieter R Roelfsema, Christian N L Olivers [2]"A self-organized neural comparator" Ludueña, G. A., & Gros, C. (2013).