Learning Goal-Conditioned Subgoals through Frozen Hierarchical World Models
Hierarchical world models plan over long horizons by passing predicted states from a coarse model to a lower-level controller.
ProofPaper ↗
Key points
- Choosing these states can be expensive: the upper-level planner repeatedly differentiates through candidate macro-action trajectories.
- We replace this search with a lightweight goal-conditioned network trained through the frozen high-level predictor.
- Inference retains an autoregressive high-level rollout, but requires no high-level candidate optimization.
- We evaluate this approach on the shared-latent HWM checkpoints distributed with H-JEPA, using three backbone seeds and three head initializations per backbone.
Sources (1)
- [1]Learning Goal-Conditioned Subgoals through Frozen Hierarchical World ModelsarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 19, 11:13 AM
Hierarchical world models plan over long horizons by passing predicted states from a coarse model to a lower-level controller.
Choosing these states can be expensive: the upper-level planner repeatedly differentiates through candidate macro-action trajectories.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 19, 2026Vector-Quantized World Model: Latent Feedback Motion Planning for Robot Manipulation
- Oct 18, 2026World Model Interpretability: From Decoding Representations to Understanding Latent Worlds
- Oct 17, 2026[AINews] High Taste Testers Ep. 1
- Oct 17, 2026Nemotron 3.5 Bolt, Muse Spark Open, and a Fast LFM2.5 24B
- Oct 16, 2026nvidia/NVIDIA-Nemotron-Labs-Dory1.0-4B-MXFP8
- Oct 16, 2026Beyond Action Prediction: Learning Search-Free Planning in Latent Space