2026-10-26 ~ 11-01 · Embodied AI
Generative · Personalization · AI Search · Infra · Embodied AI
- 1. OXE-AugE: A Large-Scale Robot Augmentation of OXE for Scaling Cross-Embodiment Policy Learning
Generalist robot policies must control many arm and gripper combinations, but recollecting data and retraining for each new robot is too costly. The widely used Open X-Embodiment dataset gathers over sixty robot…
Appeared in: 10-27
- 2. Coupling Execution Horizon and Sampling Depth in Flow-Matching VLA Policies · 📖 Deep dive
In flow-matching vision-language-action models, solver steps and execution horizon turn out to be coupled, and the best setting can reverse. These models start from random noise and use a numerical solver to push it…
Appeared in: 10-27
- 3. FLORA: An Open Model Suite for Scaling Robotics Data Engines
Training robot foundation models takes huge amounts of labels: deployments need task-success judgments, reinforcement-learning post-training needs per-step process scores, and pretraining needs timestamped captions to…
Appeared in: 10-27
- 4. OpenJaws: An Open Hardware UMI Framework with Stereo Localization for Generalist Manipulation · 📖 Deep dive
Collecting robot demonstrations in the wild with a handheld gripper (UMI) is now common, but how good the action labels are depends on one step: recovering the gripper trajectory from the wrist-camera video…
Appeared in: 10-29
- 5. Pushing Dexterity and Safety for In-Hand Object Pose Reaching That Works Remarkably Well
Skills learned by multi-fingered hands mostly stop at simple grasps or spinning an object in place. For harder in-hand moves the most common failure is dropping the object, yet earlier work rarely counts drops…
Appeared in: 10-29
- 6. Zero-HLM: Generalizable Humanoid Loco-Manipulation without Real Robot Data · 📖 Deep dive
Teaching a humanoid a long task such as walking to a table, picking up a banana and putting it in a basket usually takes a hundred or more teleoperated real-robot demonstrations…
Appeared in: 10-28
- 7. DiscoDemo: Discovering Efficient and Diverse Robot Demonstrations for Imitation Learning
Imitation learning depends heavily on how widely its demonstrations cover the task, but collecting diverse successful demonstrations is expensive. This paper uses reinforcement learning as a data generator rather than…
Appeared in: 10-27
- 8. GRAFT: Geometry-Aware Residual Adaptation for Foundation Trackers
Humanoid behavior foundation models (BFMs, general tracking policies that turn a reference motion into whole-body joint commands) are mostly trained on flat-ground motion and tend to fall on stairs, gaps and platforms…
Appeared in: 10-29
- 9. PlayPolicy: Dexterity from Supervised Play
To learn to push objects or turn an object in hand to any pose, a dexterous hand usually needs either demonstrations for each task or reinforcement learning in simulation followed by transfer to the real robot…
Appeared in: 10-28
- 10. BitVLA: 1-bit Vision-Language-Action Models for Robotics Manipulation
Vision-language-action models are too large for the edge devices robots carry. Rather than compressing after training, BitVLA uses extreme low-bit weights from the start: every parameter is one of minus one…
Appeared in: 10-27