Apple paper: generalist model AMP1 hits 85.6% zero-shot robot manipulation, beating VLAs
rohanpaul_ai · x · 2026-11-10
A new Apple paper ("On Agentic Manipulation Policies", incl. Vladlen Koltun) shows a generalist frontier model directly commanding a robot arm outperforms specialist policies like VLAs and world-action models — with zero demonstrations or task training.
Key results: the AMP1 reference harness runs unchanged across single-arm, bimanual, and mobile manipulation; it achieves 85.6% zero-shot success on RoboLab-120 vs 49.2% (leading VLA) and 42.9% (WAM); it succeeded in 104 of 108 real-world trials across 12 tasks and 3 robots, cutting failure rates 20x and 15x vs strong baselines. The harness supplies no grasping/pushing/placing skills — the agent develops its own strategies (searching for hidden objects, using objects as tools, calibrating its cameras). Adding a persistent cross-episode memory store lets it learn from experience and further cut failure rates on RoboDojo.
More from Embodied
- Developer uses a 1.9B decision model on a laptop to flag dirty restaurant tables in real time — JosephJacks_ · 2026-11-10
- Fable 5.2 beats specialized robotics AI on dual-arm RoboDojo benchmark — FinanceYF5 · 2026-11-10
- LingBot-Touch 开源触觉基础模型：近 5000 小时真实数据预训练 — dair_ai · 2026-11-10
- Norway runs 2,000+ underwater AI laser robots delousing 50 million salmon around the clock — cephaloform · 2026-11-10
- Anthropic robotics model drop: bitter lesson may hold in robotics — beffjezos · 2026-11-10
- Rider shares first-person Cybercab experience: skipping downtown driving in Austin — yunta_tsai · 2026-11-10