Loading sim environment…
One shared, crowd-trained policy. It improves in batches as the community's demos accumulate — not in real time.
Demos are accumulating. The first training run will start automatically at 500 demos.
The brain is a wrapped, fine-tuned open policy — not a foundation model. Training is batched, kicked off by operators, not by users. No real-robot transfer is claimed.
No versions trained yet.