World models of environment, agent and joint agent-environment systems
Applying computational mechanics to refine reinforcement learning through canonical predictive modeling
World models are a central component of model-based reinforcement learning. They are usually discussed in terms of what variables they predict, such as observations, rewards, states, latent or information states. We argue that there is a prior distinction: which channel they mode...
ArXiv CS.AI
Source
Generated insight by OnAirToday AI
This research provides a rigorous framework for categorizing world models by distinguishing between environment, agent, and joint system channels. By using computational mechanics to derive canonical predictive models, it reveals how closed-loop interactions can significantly reduce the complexity of agent-environment systems.
Get the day's AI news in one short email. Free.
Subscribe to our Free AI Newsletter
The Daily AI Digest, every morning. Pick your topics after you subscribe.
Free. Unsubscribe anytime.
Comments
Keep comments relevant and respectful. No links, advertising, self-promotion, or spam.