EmbodiedSmith scales embodied data via recursive self-improvement flywheel in simulation
zhenjun_zhao · x · 2026-10-08
EmbodiedSmith: a data flywheel for embodiment
EmbodiedSmith proposes scaling embodied robot data through a recursive self-improvement flywheel in simulation, targeting the scarcity and cost of real-world robot data.
The post only shares the title and a one-line tl;dr; method details and data scale require reading the paper itself.
Related event: EmbodiedSmith: Recursive Self-Improvement for Embodied Data(2 posts)→
More from Embodied
- Nature warns discreet AI wearables risk normalizing non-consensual recording; Subtle founder responds — tianshi_li · 2026-10-08
- Inside Ambarella: the low-power vision chips set to power robots, drones and AI glasses — Scobleizer · 2026-10-08
- Hierarchy-GBP accelerates factor graph inference via coarse abstraction and recovery, SOTA BA runtimes — zhenjun_zhao · 2026-10-08
- EvoMem-VLA adds state-evolution memory for long-horizon robot manipulation, hitting 80.7% on RMBench — zhenjun_zhao · 2026-10-08
- Clear Oakley Meta Vanguard smart glasses earn Z87 jobsite safety certification — armand_ruiz · 2026-10-08
- DeepMind's Raia Hadsell on continual learning, navigation and real-world robotics opportunities at RAAIS — nathanbenaich · 2026-10-08