Vision Language Actions using latent representation as action tokens
Abstract
Vision Language Action (VLA) models use latent representations as action tokens, meaning actions are not hard-coded or discretized beforehand, but learned as compact latent vectors. These latents are generated by the model (often via an autoencoder or diffusion model) and treated like tokens in a sequence, enabling unified reasoning over perception, language, and control while improving generalization and scalability across tasks and embodiments.
Fields of Study & Subjects
Robot Manipulation and Learning
Multimodal Machine Learning Applications
Reinforcement Learning in Robotics
Publication details
- Journal / source
- Open Research Knowledge Graph (ORKG)
- Publication date
- 4 February 2026
- Language
- EN
- Document type
- Dataset