Episodes

  • PoLAR: Factorizing Extent and Mode in Latent Actions for Robot Policy Learning

    PoLAR: Factorizing Extent and Mode in Latent Actions for Robot Policy Learning
    Jun 29 2026
    Introduces PoLAR, a method that factorizes latent action representations into extent and mode components to improve robot policy learning efficiency and generalization.
    Show More Show Less
    13 mins
  • Continual Robot Policy Learning via Variational Neural Dynamics

    Continual Robot Policy Learning via Variational Neural Dynamics
    Jun 29 2026
    Proposes a variational neural dynamics framework for continual robot policy learning, enabling robots to acquire new skills without forgetting previously learned ones.
    Show More Show Less
    33 mins
  • PhysisForcing: Physics-Reinforced World Models for Robotic Manipulation

    PhysisForcing: Physics-Reinforced World Models for Robotic Manipulation
    Jun 29 2026
    Plug-and-play training framework that enforces physical plausibility in robotic video generation models, achieving SOTA on R-Bench, PAI-Bench, and EZS-Bench. Lifts WorldArena success rate from 16% to 24% with zero extra inference cost.
    Show More Show Less
    24 mins
  • Translation as a Bridging Action

    Translation as a Bridging Action
    Jun 29 2026
    Replaces noisy 6DoF hand poses with relative wrist translation as a shared action space between cheap human videos and bimanual robots. Scales data-efficiently and outperforms full-pose baselines on manipulation tasks.
    Show More Show Less
    37 mins
  • Play2Perfect: Dexterous Play Pretraining for Precise Assembly

    Play2Perfect: Dexterous Play Pretraining for Precise Assembly
    Jun 28 2026
    Pre-trains a dexterous hand via unstructured 'play' interactions with objects, then fine-tunes for precise assembly tasks including 0.5 mm clearance insertions and furniture screwing, achieving 33x better sample efficiency than RL from scratch.
    Show More Show Less
    16 mins
  • Dexora: Open-Source VLA for High-DoF Bimanual Dexterity

    Dexora: Open-Source VLA for High-DoF Bimanual Dexterity
    Jun 28 2026
    First open-source Vision-Language-Action (VLA) model for dual-arm, dual-hand 36-DoF dexterous manipulation, trained on 100K simulated and 10K real trajectories with strong cross-embodiment transfer capabilities.
    Show More Show Less
    38 mins
  • WorldVLA: Towards Autoregressive Action World Model

    WorldVLA: Towards Autoregressive Action World Model
    Jun 28 2026
    Unifies VLA and world-modeling in a single autoregressive transformer that predicts both future images and actions. Outperforms separate VLA or world models on LIBERO simulation benchmarks.
    Show More Show Less
    29 mins
  • HumDex: Humanoid Dexterous Manipulation Made Easy

    HumDex: Humanoid Dexterous Manipulation Made Easy
    Jun 28 2026
    HumDex targets humanoid dexterous manipulation, aiming to simplify the development of dexterous manipulation capabilities for humanoid robots.
    Show More Show Less
    33 mins