R-AIF: Solving Sparse-Reward Robotic Tasks from Pixels with Active Inference and World Models

Re-assign community