This survey paper provides a comprehensive review of World-Action Models (WAMs) for robot learning and control. It organizes existing methods into a unified taxonomy, covering representations, transition modeling, action interfaces, architectures, training pipelines, data modalities, and scaling strategies. The paper also examines WAM applications in manipulation, navigation, and autonomous driving, while discussing key challenges such as action alignment, spatial consistency, and long-horizon memory. The goal is to establish a technical foundation for integrating predictive world modeling with action generation to enhance embodied robot intelligence. AI
IMPACT Provides a foundational technical overview for integrating predictive world modeling with action generation in robots.
RANK_REASON The item is a survey paper published on arXiv. [lever_c_demoted from research: ic=1 ai=1.0]
- action-conditioned video generation
- arXiv
- autonomous driving
- Hugging Face
- manipulation
- navigation
- reactive VLA policies
- Robot Learning and Control
- Vision-language-action policies
- World-Action Models
- World Models
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
