One-Step Prediction Fidelity Is a Poor Proxy for Planning Performance in JEPA World Models

arXiv:2610.19375 · cs.LG · Submitted 2026-10-14 · Read on arXiv

cs.LG

Submitted: 2026-10-14

Updated: 2026-10-14

Comments: 9 pages, 2 figures, 2 tables. Accepted at the NeurIPS 2026 Workshop on World Models in Physical AI (WM-PAI)

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Fidelity — 1× in this paper · explained in 2 episodes
• World Model — 1× in this paper · explained in 7 episodes

Related papers

• Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
• AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
• Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
• Convergence issues in Relational Concept Analysis based on AOC-posets
• Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks