I mean you can consider something like Dreamer, an RL agent I’ve seen. It trains a model to predict the dynamics of a system, and then trains the behavior using that model. I don’t see how this RL agent is compatible with your comment.
I mean you can consider something like Dreamer, an RL agent I’ve seen. It trains a model to predict the dynamics of a system, and then trains the behavior using that model. I don’t see how this RL agent is compatible with your comment.