Stanford's HomeBody Wires GPT-6 Astra Directly to a Unitree G1, Skips the VLA Layer
Summary
Stanford's TML lab published HomeBody, which hands a frontier VLM (GPT-6 Astra) direct control of a Unitree G1's five composable skills, bypassing the usual VLA policy layer. The robot first explores an unseen kitchen, builds a Real2Sim digital twin from its own camera/SLAM/joint data as spatial memory, then executes long-horizon cleanup and object-retrieval tasks from underspecified prompts without any environment-specific training. GPT-6 Astra runs remotely and selects skills over structured tool calls from the ego view, map, gripper state and recalled observations.
Shared on Bluesky by 3 AI experts
-
Ethan Mollick @emollick.bsky.social
It is strange how much LLMs turned out to be able to solve such a wide range of hard problems that would not, instinctively, seem to be problems that a model of human language would be able to solve This is from a Stanf…
View on Bluesky →
Originally reported by stanford.edu
Read the original article →Original headline: Stanford's HomeBody Wires GPT-6 Astra Directly to a Unitree G1, Skips the VLA Layer