robotics
The World Model Separates the Arm From What It Moves
Skeleton supervision grounds one latent-action channel in robot kinematics while a residual channel retains the rest of the visual change.
Summary
Skeleton supervision grounds one latent-action channel in robot kinematics while a residual channel retains the rest of the visual change.
Video-trained robot world models can blur together commanded motion, object response and incidental scene change. SkeLAM splits latent action into a kinematic component supervised by robot skeletons and a residual component for complementary visual dynamics. Cross-modal reconstruction encourages RGB- and skeleton-derived kinematics to align, while branch dropout limits reliance on future-derived residual information. A lightweight adapter maps real robot actions into the kinematic space at inference. On DROID, the authors report a 19.5 percent reduction in end-effector displacement error and an 11.1 percent LPIPS improvement against DreamDojo, with qualitative human-to-robot motion transfer. The findings concern the evaluated data, metrics and rollouts; generated agreement is not a safety guarantee for physical execution.
Why it matters
Skeleton supervision grounds one latent-action channel in robot kinematics while a residual channel retains the rest of the visual change.
Limits and context
- The findings concern the evaluated data, metrics and rollouts; generated agreement is not a safety guarantee for physical execution.
Key claims
Skeleton supervision grounds one latent-action channel in robot kinematics while a residual channel retains the rest of the visual change.
Qualification: The findings concern the evaluated data, metrics and rollouts; generated agreement is not a safety guarantee for physical execution.
Evidence: source-2026-10-25-009
Sources
- arXiv preprint 2610.31474, original v1arXiv · primary research
Corrections
No corrections have been recorded for this story.