PointZero: a 3D World Model Pretrained on 2.9M Synthetic Frames Without Robot Data
realChrisChoy · x · 2026-09-18
Researchers introduce PointZero, a 3D world model pretrained on 2.9M synthetic frames with zero robot data. Its core idea: learning to complete 3D point tracks yields transferable 3D dynamics, sidestepping expensive robot data for dexterous manipulation. After fine-tuning, it achieves SOTA on zero-shot 3D dynamics, action-conditioned 3D dynamics, and imitation learning, offering a new path to pretraining robot manipulation models without hardware.
Related event: PointZero trains 3D world models without robot data, hits SOTA(2 posts)→
More from Embodied
- Autonomous's Harness Pro uses an ESP32-P4, not a mini PC, to monitor AI coding agents — dee_hw · 2026-11-02
- KAIST's QCP lets robot policies control execution quality in real time — kaist-ai · 2026-11-02
- ANYbotics begins customer deliveries of ANYmal X, first legged robot certified for explosive atmospheres — lukas_m_ziegler · 2026-11-02
- RAM prices tripled, so Tesla cut a third of memory from Optimus's brain — aakashgupta · 2026-11-02
- Too lazy on weekends? Watch this robot tidy up a desktop — flngr · 2026-11-02
- Wearing Whoop and Apple Watch on a single dual-slot band — tinyfool · 2026-11-02