Search papers, labs, and topics across Lattice.
This paper introduces the CAT framework, which utilizes a trajectory-level continuous action representation for robotic manipulation, addressing the limitations of existing systems that entangle action representation with control frequency. By encoding action trajectories into continuous latent tokens and incorporating frequency-aware positional encoding, CAT achieves temporal consistency and prevents representational redundancy. Extensive evaluations show that CAT outperforms both VQ-based and continuous visuomotor baselines in long-horizon manipulation tasks, indicating its effectiveness across varying control rates and model backbones.
CAT's trajectory-level continuous action representation significantly enhances robotic manipulation success rates by eliminating representational redundancy and ensuring temporal consistency.
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.
