Patch Policy uses a frozen internet-pretrained ViT and a block-causal mask for robot control
furongh · x · 2026-07-23
A repost highlights the core idea behind Patch Policy, a robot policy method:
- A frozen internet-pretrained ViT feeds all patch tokens into a transformer-based policy head such as VQ-BeT or Diffusion Policy.
- The method adds a block-causal mask to preserve causal attention across time.
- The cited tweet frames the recipe as surprisingly simple, emphasizing reuse of a frozen vision backbone instead of training everything end-to-end.
This is a compact but technically substantive robotics-learning idea rather than a general commentary.
Related event: Patch Policy Outperforms 7B VLA with 0.7% Parameters(5 posts)→
More from Embodied
- Skyfall AI's SKY-001 world model turns sketches into CAD builds in 3.9s, vs 199s for GPT-6 Astra — rohanpaul_ai · 2026-10-27
- Tesla FSD approved in Latvia, the 10th European country; Musk praises smaller nations for moving fastest — elonmusk · 2026-10-27
- WanGP v18.81 adds Tencent's Hy-Nomas world model, playable on 8GB VRAM — cocktailpeanut · 2026-10-27
- NVIDIA open-sources FLORA, a robotics data engine ~125x cheaper than GPT-6 Astra — nvidia · 2026-10-27
- Embodied AI startups need a different operational discipline, investor says — ___Mufasaa · 2026-10-27
- π0.5 study: raising demos from 10 to 438 lifts Place Ball in Bowl success from 7% to 83% — stingning · 2026-10-27