Value-guided action planning with JEPA world models
Reading time: 2 minute
...
📝 Original Info
- Title: Value-guided action planning with JEPA world models
- ArXiv ID: 2601.00844
- Date: 2025-12-28
- Authors: Matthieu Destrade, Oumayma Bounou, Quentin Le Lidec, Jean Ponce, Yann LeCun
📝 Abstract
Building deep learning models that can reason about their environment requires capturing its underlying dynamics. Joint-Embedded Predictive Architectures (JEPA) provide a promising framework to model such dynamics by learning representations and predictors through a self-supervised prediction objective. However, their ability to support effective action planning remains limited. We propose an approach to enhance planning with JEPA world models by shaping their representation space so that the negative goal-conditioned value function for a reaching cost in a given environment is approximated by a distance (or quasi-distance) between state embeddings. We introduce a practical method to enforce this constraint during training and show that it leads to significantly improved planning performance compared to standard JEPA models on simple control tasks.📄 Full Content
World models are a class of deep learning architectures designed to capture the dynamics of systems (Ha & Schmidhuber (2018); Ding et al. (2025)). They are trained to predict future states of an environment given a sequence of actions. By explicitly modeling the system's dynamics, they capture a causal understanding of how actions influence future outcomes, enabling reasoning and planning over possible trajectories.
…(본문이 길어 일부가 생략되었습니다.)
Reference
This content is AI-processed based on open access ArXiv data.