Aim Short to Reach Far: Your Frozen World Model Can Plan Better Than You Think
cs.LG, cs.RO
Submitted: 2026-09-24
Updated: 2026-09-30
License: http://creativecommons.org/licenses/by/4.0/
The gist: Latent world models plan toward goal images with a frozen pretrained predictor, without task rewards or extra trained heads.
Terminology
Abstract
Latent world models plan toward goal images with a frozen pretrained predictor, without task rewards or extra trained heads. However, their planners struggle with long-range goals, and prior work addresses this by training extra components such as value functions or subgoal models. We show that the planning target itself can cause this failure: even with exact dynamics and globally optimal short-horizon search, scoring predictions by their distance to the final goal rejects the first steps of a route that initially moves away from the goal. Building on this insight, we propose Anchored Planning (AP), a training-free method that reuses the world model's own offline trajectories. AP retrieves a segment that leads from the current observation toward the goal and aims the frozen planner at an observation shortly after the segment's start. Across four diverse tasks, AP substantially improves frozen LeWM planners for both action synthesis and action ranking, and it outperforms both additional final-goal search and the LeWM planner on long-range goals.
Sources
- V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning
- LeJEPA: Provable and Scalable Self-Supervised Learning Without the Heuristics
- Mind the Gap: Promises and Pitfalls of Hierarchical Planning in LeWorldModel
- Latent World Models with Monotone Planning Costs for Image-Goal Navigation
- SAGE: Subgoal-Conditioned Action Generation for Latent World Model Planning
- Visual Foresight: Model-Based Deep Reinforcement Learning for Vision-Based Robotic Control
- IMWM: Intuition Models Complement World Models for Latent Planning
- TD-MPC2: Scalable, Robust World Models for Continuous Control
- LeWorldModel: Stable End-to-End Joint-Embedding Predictive Architecture from Pixels
- The Objective Is the Bottleneck: Latent World Models Encode What Their Planners Cannot Use
- Hierarchical Planning with Latent World Models
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