arXiv · 2611.22470
Escaping Latent Traps by Remembering: Training-Free Episodic Kernel Repulsion for World-Model Planning
Abstract
Latent world models plan towards a goal image by minimising the distance between predicted and goal latents. This distance is not geodesic: it has local minima, latent traps, behind walls, at corners and in dead ends, and memoryless receding-horizon replanning keeps the agent in them, revisiting states far from the goal. We propose Episodic Kernel Repulsion (EKR), a training-free cost term that penalises candidate plans by a top-$k$ Gaussian kernel to a novelty-gated memory of the latents the agent has actually visited, rescaled by the current base cost so that its strength is comparable across models. Across two JEPA world models and four benchmarks, EKR raises average success from 31.8% to 46.7% (LeWM) and from 54.7% to 64.0% (H-JEPA) with the same model, planner and budget. Same-memory counterfactuals show that, in most settings, EKR pushes plans away from visited states, in navigation, often at a higher base cost; stuck episodes and revisits drop accordingly. In seven of eight settings it matches or exceeds $4\times$ samples at $0.8$--$1.5\times$ compute. Without tuning, it transfers to DINO-WM with CEM: $+6.7$ and $+19.3$ points on official and longer-goal Wall tasks, and from 44.7% to 100.0% and from 21.0% to 32.0% on long-horizon PointMaze and Push-T.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Yican Zhao, Bolin Yang, Ruizhe Wang, Haiyang Guo, Fuji Yang, Bingo Wing-Kuen Ling, Zelin Zang. 2026-11-17. Escaping Latent Traps by Remembering: Training-Free Episodic Kernel Repulsion for World-Model Planning. https://arxiv.science/abs/2611.22470
Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite arXiv Science by Indiciae: BibTeX · RIS.