aiqbench analysis
This paper introduces CAT, a novel framework that decouples action representation from control frequency in robotic manipulation by encoding trajectories into continuous latent tokens. By incorporating frequency-aware positional encoding and trajectory-level regularization, the method eliminates representational redundancy at high sampling rates while maintaining temporal consistency across varying control rates. System evaluations on both simulated and real-world long-horizon tasks demonstrate that CAT consistently outperforms existing VQ-based and continuous visuomotor baselines in success rates.
Trajectory-Level Continuous Action Representation for Robotic Manipulation
arXiv:2608.24111v1 Announce Type: new Abstract: We propose CAT, a trajectory-level continuous action representation framework for robotic manipulation. Existing visuomotor systems often entangle action representation with control frequency or rely on fixed temporal parameterizations. This leads to representational redundancy at high sampling rates and limits the modeling of critical motion. CAT instead encodes action trajectories within a fixed real-time interval into a set of continuous latent tokens. To ensure temporal consistency across varying control frequencies, we further incorporate a frequency-aware positional encoding that establishs a shared temporal coordinate system. Trajectory-level regularization further stabilizes the latent representation. This approach prevents representation growth with timestep density and avoids reliance on predefined temporal parameterizations. Extensive system-level evaluations on LIBERO, MimicGen, and real-world long-horizon manipulation tasks demonstrate that CAT-based policies consistently outperform both competitive VQ-based and continuous visuomotor baselines under matched training settings. Across various model backbones and control frequencies, CAT consistently improves success rates. These results highlight the advantages of trajectory-level continuous action modeling for scalable robotic manipulation across varying control rates.
Abstract:We propose CAT, a trajectory-level continuous action representation framework for robotic manipulation. Existing visuomotor systems often entangle action representation with control frequency or rely on fixed temporal parameterizations. This leads to representational redundancy at high sampling rates and limits the modeling of critical motion. CAT instead encodes action trajectories within a fixed real-time interval into a set of continuous latent tokens. To ensure temporal consistency across varying control frequencies, we further incorporate a frequency-aware positional encoding that establishs a shared temporal coordinate system. Trajectory-level regularization further stabilizes the latent representation. This approach prevents representation growth with timestep density and avoids reliance on predefined temporal parameterizations. Extensive system-level evaluations on LIBERO, MimicGen, and real-world long-horizon manipulation tasks demonstrate that CAT-based policies consistently outperform both competitive VQ-based and continuous visuomotor baselines under matched training settings. Across various model backbones and control frequencies, CAT consistently improves success rates. These results highlight the advantages of trajectory-level continuous action modeling for scalable robotic manipulation across varying control rates.
Robotics (cs.RO)
arXiv:2608.24111 [cs.RO]
(or arXiv:2608.24111v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2608.24111 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 256 words
Related coverage
Building a Foundation Stack for General-Purpose Robots
IEEE Spectrum · Jul 13, 2026
GEAR-VLA: Learning Geometry-Aware Action Representations for Generalizable Robotic Manipulation
arXiv · Jun 11, 2026
GEAR-VLA: Learning Geometry-Aware Action Representations for Generalizable Robotic Manipulation
arXiv · Jun 9, 2026
Dyna-2: General Purpose Physical Agents with Hour-Long Autonomy — Jason Ma, Dyna Robotics
Latent Space · Oct 13, 2026