Causal World Modeling for Robot Control

Lin Li, Qihang Zhang, Yiming Luo +9

This work highlights that video world modeling, alongside vision-language pre-training, establishes a fresh and independent foundation for robot learning. Intuitively, video world…