# RL Seminar - Final Project ## World Model Training ### Action-conditioned generation MiniGrid-RedBlueDoors ![](https://i.imgur.com/CRs4a0r.gif =500x) ![](https://i.imgur.com/869ouwj.gif =500x) ![](https://i.imgur.com/9Lgouoa.gif =500x) DMLab - Watermaze ![](https://i.imgur.com/jtjuUFM.gif =500x) ![](https://i.imgur.com/jTxceLz.gif =500x) ![](https://i.imgur.com/vz6bTqs.gif =500x) ![](https://i.imgur.com/tEbIc9X.gif =500x) ### Predict rewards and Termination ![](https://i.imgur.com/6G4Z4AS.gif) ![](https://i.imgur.com/KPx553Q.gif) ![](https://i.imgur.com/4pRZoyZ.gif) ![](https://i.imgur.com/T5VpNKt.png) ![](https://i.imgur.com/IruY5im.png) ![](https://i.imgur.com/3sn8qGb.png) ![](https://i.imgur.com/CJ8H3s7.png =200x) ![](https://i.imgur.com/KhjBJcZ.png =200x) ![](https://i.imgur.com/XU1sLGv.png =200x) ## Reinforcement Learning Training ![](https://i.imgur.com/OJGBnVk.png)