Fetching the paper…

Backprop-Free Reinforcement Learning with Active Neural Generative Coding · Around