Search papers, labs, and topics across Lattice.
This paper introduces ACID, a novel decision-time planning framework that enhances action-conditioned world models by incorporating cycle action consistency, ensuring that actions inferred from predicted transitions align with those originally conditioned on. The significance of this approach lies in its ability to maintain the realizability of intermediate transitions, addressing a critical shortcoming in traditional planning methods. Experimental results across various tasks demonstrate that ACID not only improves planning efficiency but also achieves comparable accuracy to baseline methods while requiring significantly less computational resources.
Cycle action consistency in planning can drastically reduce computational costs while maintaining accuracy across diverse manipulation and navigation tasks.
Decision-time planning with action-conditioned world models has become a popular paradigm for embodied control. However, the standard planning cost judges a candidate solely by how close its predicted terminal state lies to the goal, leaving the realizability of the intermediate transitions unchecked -- a predicted trajectory can look convincing while the environment rollout drifts away from it. In this paper, we propose ACID, a decision-time planning framework that introduces cycle action consistency: the action inferred backward from a predicted transition by an inverse dynamics model should recover the one that was conditioned on. We fold this per-step residual into the planning cost via a scale-invariant adaptive weight. Across four action-conditioned world models and six tasks spanning rigid and deformable manipulation, articulated control, and visual navigation, ACID consistently improves planning and matches the baseline's accuracy with substantially less planning compute.
