New paper benchmarks pi vs GR00T: how many demos do VLAs need to learn a task?
rahulshivchand · x · 2026-11-11
A new paper studies sample efficiency across vision-language-action models: with enough demos VLAs achieve high success rates, but how many demos and training steps does a new task actually take? The authors benchmark pi and GR00T to find out, with a detailed thread.
More from Embodied
- Beni robot, #1 personal robot on Kickstarter, officially launches with self-recovery and 4K60 — Scobleizer · 2026-11-11
- Berkeley's Social Humanoid teaches Unitree G1 socially responsive interaction — berkeley_ai · 2026-11-11
- Robotics is not solved: frontier models are ~6 orders of magnitude off real assembly lines — ZeeshanZiaML · 2026-11-11
- MacBooks do have a hinge angle sensor—verified with Codex, prompt included — lipeng0820 · 2026-11-11
- DynaRobotics robot demo at CoRL sparks 'I want one at home' reactions — JasonMa2020 · 2026-11-11
- Humanoid Factory Deploy Stalls: Teleop Intervention Every 6 Minutes After 15 Weeks — ZeeshanZiaML · 2026-11-11