Transformer Transformer: feed an UMI dataset, get an optimized mobile manipulator
SongShuran · x · 2026-09-06
haqhuy announces "Transformer Transformer": conditioning on diverse motions (e.g. an UMI dataset), the method co-optimizes a mobile manipulator for them — data-driven robot body+policy co-design. The author notes sundayrobotics' Memo torso design proved critical, thanking collaborators. Accepted at CoRL 2026.
Related event: Stanford's Transformer Transformer Auto-Designs Robot Morphologies(3 posts)→
More from Embodied
- HKUST(GZ) lab lands 3 CoRL 2026 papers, unveils terrain-adaptive robot motion model — Scobleizer · 2026-09-06
- World Labs unveils Atlas, an omni world model for spatial intelligence — pstAsiatech · 2026-09-06
- Musk signals Tesla could turn Cybercab tech into a consumer AI-driven car lineup — elonmusk · 2026-09-06
- Roboticist: Next Frontier Is Truly Leveraging AI Models for Manipulation, Not Agent Wrappers — chris_j_paxton · 2026-09-06
- Altman says AI may create an entirely new computer category built around proactivity — rohanpaul_ai · 2026-09-06
- Cybercab ships with a privacy mode, and whurley thinks it's a standout feature — whurley · 2026-09-06