ACT-2 Hits 99.1% Generalization in Laundry Folding
tonyzzhao · x · 2026-07-17
ACT-2 underwent one of its most rigorous generalization tests for laundry folding tasks.
The results show:
- 785 trials completed across 31 unseen environments
- A success rate of 99.1%
- Folding quality approaching human levels, with an average score of 4.72/5
- 98.3% of samples received 4 or 5 stars
The key takeaway isn't just a single successful demo, but the system's stability in novel environments and its near-human folding quality.
Related event: Sunday Robotics Unveils ACT-2 Preview for Home Robots(17 posts)→
More from Embodied
- Full teardown: inside Zmotor's 32mm linear actuator for humanoids — MarwaEldiwiny · 2026-10-27
- Full Teardown of a Zmotor Linear Actuator Used in China's Humanoid Robots — MarwaEldiwiny · 2026-10-27
- Video AI agent runs a physics engine, then paints your real room over the blocks — jonstephens85 · 2026-10-27
- World Model Assigns Weight and Friction to Objects, Then Simulates with a Physics Engine — jonstephens85 · 2026-10-27
- First unboxing shots of Apple's foldable "iPhone Duo" surface — chrisfirst · 2026-10-27
- UniAction: unified action conditioning lets one world model handle mixed action types — ZeYanjie · 2026-10-27