OneWM: unified causal training shows world and action modeling improve each other
IAAI233 · hf · 2026-10-16
Researchers from HKU MMLab, SJTU, THU and others propose OneWM, a framework that jointly trains world and action modeling under shared causal training, treating actions as both conditions for world prediction and targets for policy learning. Experiments on LIBERO, LIBERO-Plus and RoboCasa365 plus real-world tests show learning action-conditioned dynamics improves policy performance and generalization, while action learning in turn improves multi-view world prediction.
More from Embodied
- Berkeley researchers sample humanoid robot policies directly from weight space with diffusion — berkeley_ai · 2026-10-16
- Tesla's hover car isn't a flying car — compressed air can't sustain flight — tinyfool · 2026-10-16
- Scobleizer relays claim that Optimus demo will be even more mind-blowing than the road-flipping Roadster — Scobleizer · 2026-10-16
- Indie dev's AI PASSPORT wearable firmware hits 1,200+ downloads in days — ezshine · 2026-10-16
- 1X Opens California Factory for Mass Production of NEO Humanoid Robots — emmanuelvivier · 2026-10-16
- Waymo Fully Opens Robotaxi Service in San Antonio and Two Other Cities — emmanuelvivier · 2026-10-16