Weekly Top 10 · W47

2026-11-16 ~ 11-22 · Embodied AI

Generative · Personalization · AI Search · Infra · Embodied AI

  1. 1. What Makes a VLA Reactive? A Study of Execution Strategies for Dynamic Manipulation · 📖 Deep dive

    A VLA model predicts a short chunk of future actions at a time, and when objects move, that chunk goes stale before it finishes executing. The authors use a pretrained π0.5 on two dynamic simulation benchmarks and…

    Appeared in: 11-16

  2. 2. Scaling VLA Serving Under Robotic Execution Constraints · 📖 Deep dive

    The authors' company already has hundreds of VLA-driven robots working in real environments, and robots must share GPUs to keep inference costs down. Robots keep executing old actions while waiting for new ones…

    Appeared in: 11-16

  3. 3. Learning Low-Latency Action Verification for Frozen Vision-Language-Action Models

    The actions a VLA model samples are not always good, and sampling several and picking one is safer, but using a large model as the judge adds a long wait to every query. The authors' HDH trains its judge only on…

    Appeared in: 11-16

  4. 4. Training Reasoning Models in Simulation to Natively Use Real Robots

    Open-weight reasoning models asked to control a real robot usually still lag far behind frontier proprietary models, and they often need demonstrations or an external harness…

    Appeared in: 11-17

  5. 5. ILA: A Transferable Interface-Level Adapter for Compliant Humanoid Motion Tracking

    Humanoid robots that track human motion are often used for teleoperation and motion data collection. But following the reference motion rigidly is unsafe in tight spaces or when a person is in physical contact…

    Appeared in: 11-16

  6. 6. X-Nav Technical Report: Real-World Long-Horizon Indoor-to-Outdoor Navigation on a Humanoid Robot

    A bipedal humanoid walking from indoors to outdoors on a long language-instructed mission passes rooms, doors, and roads, and the camera moves with the gait; few systems have actually finished such a route…

    Appeared in: 11-17

  7. 7. SALA: Semantically Aligned Latent Actions Enable Dual-Axis Scaling of Embodied Foundation Models

    Robot interaction data is expensive, while first-person human video is cheap and plentiful, and a common approach uses latent action models to extract action signals from video to supervise policies…

    Appeared in: 11-16

  8. 8. Reusing Locomotion Training Recipes Across Humanoids via Body-Aligned Virtual Joints

    Humanoid walking policies often output native joint angles, and rewards and action limits are written in those robot-specific coordinates, so a recipe tuned on one robot is hard to reuse on another…

    Appeared in: 11-17

  9. 9. FORGE: Force-Aware Multi-Granularity Control for Efficient Humanoid Loco-Manipulation

    Getting a humanoid robot to walk while carrying things makes teleoperated data collection hard: the controller must be precise, stay balanced, and know how much force its hands feel…

    Appeared in: 11-16

  10. 10. I See What You Want: Decoding Hierarchical Operator Intent from Visual Teleoperation History

    When a person teleoperates a robot, can a large model tell what they want just from the video? The authors split this into two levels, the action happening now and the overall task it serves…

    Appeared in: 11-16

‹ Previous week