How can HTM be used in Reinforcement Learning projects?

[kayoj]

Hello

I have been learning about reinforcement learning & started to wonder if Hierarchical Temporal Memory (HTM) ideas can fit into this field. Most RL projects today use deep neural networks but HTM focuses more on patterns and sequences. I think it could help with predicting states but I am not sure how well it works with reward-driven systems.

Has anyone here tried using HTM together with RL? For eg; maybe using HTM to process inputs before sending them to an RL model / to spot unusual patterns that could change how rewards are given. I would like to hear if people have tested this in real projects and what the results were like.

I checked the HTM School resources but they don’t explain much about RL. While studying this; I also learned what is a Prompt Engineer?; which shows how new roles are appearing in machine learning. If anyone has guides, examples / experiments that connect HTM and RL; that would really help.

Thank you !!

[dmac]

Hello,

Here is a project from many years ago. I hope this helps

HTM Based Autonomous Agent

Hi all, Here is the preprint of my Ms thesis with the title: “Hierarchical Temporal Memory Based Autonomous Agent For Partially Observable Video Game Environments” I defended it a couple of days ago. It proposes a non-player character architecture that combines HTM and TD(lambda) and presents results on a 3D video game learning task. The architecture incorporates abstracted functionalities of layer 5 and basal ganglia pathways in order to produce voluntary actions. It does so by presenting th…