{"slug": "end-to-end-learning-of-hierarchical-world-models-for-visual-planning", "title": "End-to-End Learning of Hierarchical World Models for Visual Planning", "summary": "A paper submitted to arXiv on 5 October 2026 introduces H-JEPA, an end-to-end recipe for training a hierarchy of action-conditioned JEPA world models in which each level predicts farther ahead in its own learned latent space. On the Visual AntMaze benchmark, a three-level H-JEPA hierarchy raised planning success from 18% to 73% while using less planner compute than a flat JEPA, and ablations attribute the gains to both temporal decomposition and higher-level goal representations. With inverse-dynamics supervision, the approach extended to real-robot videos from DROID, where the hierarchy improved offline planning fidelity at lower planner compute.", "body_md": "# Computer Science > Machine Learning\n\n  [Submitted on 5 Oct 2026]\n\n# Title:H-JEPA: End-to-End Learning of Hierarchical World Models for Visual Planning\n\n[View PDF](https://arxiv.org/pdf/2610.06805)\n\n[HTML (experimental)](https://arxiv.org/html/2610.06805v1)\n\nAbstract:Long-horizon planning with latent world models requires reasoning across timescales and levels of abstraction. Existing task-agnostic JEPA world models predict and plan at a single timescale or with multiple horizons in one shared latent space. We introduce H-JEPA, an end-to-end recipe for training a hierarchy of action-conditioned JEPAs in which each level predicts farther ahead in its own learned latent space. Planning proceeds top-down: the top level optimizes progress toward the goal, and each level's predictions become subgoals for the planner below it. When factors in the data evolve at separated timescales, higher levels discard fast, unpredictable detail and retain slower task-relevant state. Across four simulated navigation and manipulation environments, hierarchical planning improves over a flat JEPA; on Visual AntMaze, a three-level hierarchy raises success from 18% to 73% using less planner compute. Ablations attribute these gains to both temporal decomposition and higher-level goal representations. With inverse-dynamics supervision, the approach extends to diverse real-robot videos from DROID, where hierarchy improves offline planning fidelity at lower planner compute.\n    \n\n### References & Citations\n\nLoading...\n\n# Bibliographic and Citation Tools\n\nBibliographic Explorer \n\n*(*[What is the Explorer?](https://info.arxiv.org/labs/showcase.html#arxiv-bibliographic-explorer))\nConnected Papers \n\n*(*[What is Connected Papers?](https://www.connectedpapers.com/about))\nLitmaps \n\n*(*[What is Litmaps?](https://www.litmaps.co/))\nscite Smart Citations \n\n*(*[What are Smart Citations?](https://www.scite.ai/))\n# Code, Data and Media Associated with this Article\n\nalphaXiv \n\n*(*[What is alphaXiv?](https://alphaxiv.org/))\nCatalyzeX Code Finder for Papers \n\n*(*[What is CatalyzeX?](https://www.catalyzex.com))\nDagsHub \n\n*(*[What is DagsHub?](https://dagshub.com/))\nGotit.pub \n\n*(*[What is GotitPub?](http://gotit.pub/faq))\nHugging Face \n\n*(*[What is Huggingface?](https://huggingface.co/huggingface))\nScienceCast \n\n*(*[What is ScienceCast?](https://sciencecast.org/welcome))\n# Demos\n\n# Recommenders and Search Tools\n\nInfluence Flower \n\n*(*[What are Influence Flowers?](https://influencemap.cmlab.dev/))\nCORE Recommender \n\n*(*[What is CORE?](https://core.ac.uk/services/recommender))\nIArxiv Recommender\n\n*(*[What is IArxiv?](https://iarxiv.org/about))\n# arXivLabs: experimental projects with community collaborators\n\narXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.\n\nBoth individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.\n\nHave an idea for a project that will add value for arXiv's community? [**Learn more about arXivLabs**](https://info.arxiv.org/labs/index.html).", "url": "https://wpnews.pro/news/end-to-end-learning-of-hierarchical-world-models-for-visual-planning", "canonical_source": "https://arxiv.org/abs/2610.06805", "published_at": "2026-11-01 15:12:32+00:00", "updated_at": "2026-11-01 15:19:51.490817+00:00", "lang": "en", "topics": ["machine-learning", "ai-research", "robotics", "artificial-intelligence"], "entities": ["H-JEPA", "JEPA", "Visual AntMaze", "DROID", "arXiv"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/end-to-end-learning-of-hierarchical-world-models-for-visual-planning", "markdown": "https://wpnews.pro/news/end-to-end-learning-of-hierarchical-world-models-for-visual-planning.md", "text": "https://wpnews.pro/news/end-to-end-learning-of-hierarchical-world-models-for-visual-planning.txt", "jsonld": "https://wpnews.pro/news/end-to-end-learning-of-hierarchical-world-models-for-visual-planning.jsonld"}}