README_notes.md

Things to understand regarding p2e

• Whats is?:
  • reward driven planner?
  • curiosity driven planner
  • Dreamer system vs p2e system --> understand the differences

Training vs. Testing?

• Is there a difference of zero-shot learning in RL domain vs. supervised classification domain?
• see "Survey of Zero-shot Generalisation in Deep RL": https://www.jair.org/index.php/jair/article/view/14174/26890

How is the world model used?

- How it is used during training?