Hierarchical Planning

Follow this topic Ask about this topic

Hierarchical planning studies how to break complex tasks into subgoals at multiple levels so agents can plan farther ahead without searching every possible action sequence. Methods include symbolic Hierarchical Task Networks, reusable skills or options, learned latent subgoals and world models, and language models that propose or refine plans. Researchers evaluate these approaches in settings such as robot navigation and manipulation, household and web agents, multi-robot coordination, and scheduling. A central challenge is choosing useful levels of detail while ensuring abstract plans remain feasible in the real environment and can recover when execution goes wrong.

Latest papers

35 papers7 papers in the last four weeks, up 40% on the four weeks before. 0.0% of all new papers.

Hierarchical Planning with Latent World Models
Apr 3, 2026Wancong Zhang, Basile Terver, Artem Zholus +7Hierarchical Planning Hierarchical Robot Control
FAIR at Meta · New York University · Mila - Québec AI Institute +1
LDHP: Library-Driven Hierarchical Planning for Non-prehensile Dexterous Manipulation
Mar 14, 2026Tierui He, Jiahui Zuo, Fumin Zhang +1Hierarchical Planning Contact-Rich Manipulation
Shanghai Marine Equipment Research Institute, Shanghai 200031, China. · J. Zuo and F. Zhang are with The Hong Kong University of Science and Technology, Clear Water Bay, Hong Kong. · C. Zhao is with Jilin University, Changchun, China.
Bilevel Planning with Learned Symbolic Abstractions from Interaction Data
Mar 9, 2026Fatih Dogangun, Burcu Kilic, Serdar Bahar +1Hierarchical Planning Symbolic Planning
Department of Computer Engineering, Bogazici University, Istanbul, Türkiye. · Department of Computer Engineering, Middle East Technical University, Ankara, Türkiye.
Dynamic Targeting of Satellite Observations Using Supplemental Geostationary Satellite Data and Hierarchical Planning
Mar 5, 2026Akseli Kangaslahti, Itai Zilberstein, Alberto Candela +1Remote Sensing Long-Horizon Planning
Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA · Harvard University, Cambridge, MA · Department of Computer Science, Carnegie Mellon University, Pittsburgh, PA
City Editing: Hierarchical Agentic Execution for Dependency-Aware Urban Geospatial Modification
Feb 22, 2026Rui Liu, Steven Jige Quan, Zhong-Ren Peng +6Hierarchical Planning Geospatial Reasoning
University of Kansas · Seoul National University · University of Florida +3

Apr 3, 2026J/K move · Enter open · S save

Hierarchical Planning with Latent World Models

Wancong Zhang, Basile Terver, Artem Zholus, Soham Chitnis+6

FAIR at Meta · New York University · Mila - Québec AI Institute+1

Hierarchical Planning Hierarchical Robot Control Hierarchical World Models Long-Horizon Robotic Manipulation Model Predictive Control Latent World Models

Open paper AskPDF arXiv

World models are a promising path to zero-shot embodied control through planning. However, existing world model planners struggle on long-horizon, multi-stage tasks: prediction errors compound and naive search is exponential in the planning horizon. Hierarchy mitigates both by decomposing tasks into shorter, tractable subproblems; yet prior hierarchical approaches either amortize control into task-specific policies (hierarchical RL) or assume low-dimensional states and known dynamics (classical hierarchical MPC). We present Hierarchical Planning with Latent World Models (HWM), an architecture and planning paradigm for hierarchical model predictive control directly on visual world models trained solely via next-latent prediction. HWM learns world models at multiple temporal scales within a shared latent space, so predictions from the long horizon model serve as subgoals for the short horizon model via latent matching, without task-specific rewards or hierarchical policies. To keep long-horizon search tractable, HWM learns an action encoder that compresses primitive action chunks into latent macro-actions. On real-world Franka manipulation, HWM solves pick-and-place from a single goal image at 70% success vs. 0% for single-level planning. Across simulated push manipulation and maze navigation, HWM consistently improves performance on long-horizon tasks while requiring up to 3x less planning compute.

Latest papers

Hierarchical Planning with Latent World Models

LDHP: Library-Driven Hierarchical Planning for Non-prehensile Dexterous Manipulation

Bilevel Planning with Learned Symbolic Abstractions from Interaction Data

Dynamic Targeting of Satellite Observations Using Supplemental Geostationary Satellite Data and Hierarchical Planning

City Editing: Hierarchical Agentic Execution for Dependency-Aware Urban Geospatial Modification