VLA-JEPA: Enhancing Vision-Language-Action Model with Latent World Model
- 11 September 2026
- book chapter
- Published by Springer Nature
- p. 478-497
Abstract
No abstract availableKeywords
All Related Versions
This publication has 25 references indexed in Scilit:
- Moto: Latent Motion Token as the Bridging Language for Learning Robot Manipulation from Videos2025
- DyWA: Dynamics-Adaptive World Action Model for Generalizable Non-Prehensile Manipulation2025
- RoboBrain: A Unified Brain Model for Robotic Manipulation from Abstract to Concrete2025
- CoT-VLA: Visual Chain-of-Thought Reasoning for Vision-Language-Action Models2025
- ShapeLLM: Universal 3D Object Understanding for Embodied Interaction2024
- RT-H: Action Hierarchies using Language2024
- Universal Manipulation Interface: In-The-Wild Robot Teaching Without In-The-Wild Robots2024
- Self-Supervised Learning from Images with a Joint-Embedding Predictive Architecture2023
- Bridge Data: Boosting Generalization of Robotic Skills with Cross-Domain Datasets2022
- The “Something Something” Video Database for Learning and Evaluating Visual Common Sense2017