HKUST-GZ's UniWAM Unifies Physical Reasoning, World Generation and Action Prediction, Finds Co-Training Scaling Law
HKUSTGZ · hf · 2026-10-08
HKUST (Guangzhou) introduces UniWAM, a unified world-action model that jointly learns semantic understanding of the physical world, visual generation, and action prediction.
- Design: integrates a physical reasoner, world generator, and action predictor; low-level actions are represented in natural language, with complementary supervision from VQA data, human egocentric data, and robot demos to preserve language capabilities.
- Training: rigorous data cleaning pipeline; future visual noise augmentation and history-conditioned flow matching cut denoising steps while keeping performance.
- Results: SOTA across in-distribution performance, robustness, generalization, instruction following, and long-horizon tasks; uncovers a log-linear scaling law for human-robot co-training.
More from Embodied
- Developer uses Codex for CAD and build docs in physical robot hardware project — OpenAIDevs · 2026-10-08
- Twitter user dance-battles a Unitree robot in real life — zealcaiden · 2026-10-08
- Nvidia's DreamDojo world model wins ICML spotlight despite critical bugs in released code — Amazing-Fox-7295 · 2026-10-08
- Vacuum robot caught running away from home, poster says they get it — BotLordBotflix · 2026-10-08
- Bulkhead drone motor design concept targets 60k motors per month in production — MatthewChang · 2026-10-08
- This robot vacuum ran away from home and honestly, we get it — DankestMage99 · 2026-10-08