GPT6 as robot policy: end-to-end ee-frame actions from dual RGB-D, near-perfect box stacking

ChongZzZhang · x · 2026-09-21

ChongZzZhang demos "GPT6 as policy" driving a robot end-to-end: the model outputs ee-frame actions directly from two RGB-D camera streams and natural-language instructions. Tasks include finding an occluded wooden box, placing it on a yellow marker, and stacking boxes; the author estimates near-100% success with enough inference tokens and retries.

Related event: GPT6 as Policy: Wrist-Mounted Dual Cameras Drive Zero-Shot Robot Stacking(5 posts)→

Original post →

More from Embodied

Embodied channel →