Ctrl-World（ICLR 2026）

Ctrl-World is designed for policy-in-the-loop rollouts with generalist robot policies.

All videos are generated by passing in the initial frame and a different sequences of actions as input.

It generates joint multi-view predictions (including wrist views), enforces fine-grained action control via frame-level conditioning,

enforces fine-grained action control via frame-level conditioning

Ctrl-World is an action-conditioned world model compatible with modern VLA policies and enables policy-in-the-loop rollouts entirely in imagination

The model can evaluate a policy’s instruction-following capability but also can be imprecise in modeling physical interactions.

This repo include the Ctrl-World model checkpoint trained on opensourced DROID dataset (~95k trajectories, 564 scenes).