Skip to main content
Search

# Robotics World Model

1 related articles

Subscribe to updates
  1. arXiv

    Robot world models learn to notice small action changes: policy-success correlation on DROID rises from 0.47 to 0.79

    Robot world models can generate realistic manipulation video yet often ignore small action changes, predicting the same outcome for actions with different real consequences. AFFECT injects logged perturbations into training actions and requires generated videos to respond to those perturbations as recorded videos do, adding action-sensitivity supervision without architectural changes or repeated executions from the same initial state. On the DROID real-robot configuration, the Pearson correlation between policy success rates inside the DreamDojo world model and on the real robot rises from 0.47 to 0.79, with FVD lower by up to 23%, and planning with the trained model improves task success.

You’re all caught up