Vision Language Action using affordance as action tokens
Abstract
Vision Language Action using affordances as action tokens represents actions as object–action possibilities (e.g., graspable, pushable, pourable) rather than explicit motor commands. By predicting affordances grounded in visual and linguistic context, these models abstract action selection from low-level control, enabling better generalization, interpretability, and transfer across tasks and embodiments.
Fields of Study & Subjects
Robot Manipulation and Learning
Embodied and Extended Cognition
Multimodal Machine Learning Applications
Publication details
- Journal / source
- Open Research Knowledge Graph (ORKG)
- Publication date
- 2 February 2026
- Language
- EN
- Document type
- Dataset