mlsanityData to 7 Oct 2026

IMWM: Intuition Models Complement World Models for Latent Planning

arXiv 2606.01626 ↗First posted 1 Jun 2026cs.LGcited by 3

On the map: World models → Latent predictive world models

Abstract (arXiv)

Planning with a learned latent world model is a promising route to control from raw pixels, but a strong world model alone is not enough. We show this experimentally: even with a perfect world model (operationalized by replacing the learned forward predictor with an idealized rollout of the true environment dynamics), a finite-budget sample-based planner still fails on some tasks, indicating that the bottleneck can lie in search rather than in world-model accuracy. Motivated by this gap, we propose IMWM (Intuition Model + World Model), which pairs the world model with an intuition model trained from demonstrations to recognize promising actions. The two models collaborate through three lightweight components: (i) Retrieval Initialization, which initializes the planner's action proposal from a retrieved demonstration; (ii) Hybrid Cost, which combines the intuition score with the world-model rollout cost; and (iii) a Reliability Gate, which adjusts how much the planner trusts intuition in each setting. Across four pixel-based goal-reaching tasks (Two-Room, Reacher, Push-T, and OGBench-Cube), IMWM has higher mean success than the world-model-only planner on all four, with the largest gains on Two-Room (99.2%, +11.5 percentage points) and OGBench-Cube (94.7%, +28.5 percentage points).

Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Placed in Latent predictive world models, one of 11 directions in World models. Builds on 32 of its 42 references that are in the map. First posted on arXiv 1 Jun 2026.

Builds on

32 of 42 references.

Closest ancestors · 3

Same direction, newest first

  1. 2026LeWorldModel: Stable End-to-End Joint-Embedding Predictive Architecture from Pixels13 Mar
  2. 2022Temporal Difference Learning for Model Predictive Control9 Mar
  3. 2018Learning Latent Dynamics for Planning from Pixels12 Nov

Other references · 29

  1. 2018Representation Learning with Contrastive Predictive Codingcited by 7,662
  2. 2019Dream to Control: Learning Behaviors by Latent Imaginationcited by 1,598
  3. 2017Curiosity-driven Exploration by Self-supervised Predictioncited by 1,173
  4. 2022Planning with Diffusion for Flexible Behavior Synthesiscited by 1,023
  5. 2020Mastering Atari with Discrete World Modelscited by 990
  6. 2018Deep Reinforcement Learning in a Handful of Trials using Probabilistic Dynamics Modelscited by 748
  7. 2022R3M: A Universal Visual Representation for Robot Manipulationcited by 678
  8. 2022Contrastive Learning as Goal-Conditioned Reinforcement Learningcited by 241
  9. 2018Visual Foresight: Model-Based Deep Reinforcement Learning for Vision-Based Robotic Controlcited by 220
  10. 2020Accelerating Reinforcement Learning with Learned Skill Priorscited by 220
  11. 2019Deep Dynamics Models for Learning Dexterous Manipulationcited by 200
  12. 2017Neural Episodic Controlcited by 148
17 more
  1. 2017Imagination-Augmented Agents for Deep Reinforcement Learningcited by 137
  2. 2020Parrot: Data-Driven Behavioral Priors for Reinforcement Learningcited by 136
  3. 2017Value Prediction Networkcited by 117
  4. 2016Model-Free Episodic Controlcited by 100
  5. 2016The Predictron: End-To-End Learning and Planningcited by 92
  6. 2020Sample-efficient Cross-Entropy Method for Real-time Planningcited by 90
  7. 2020Objective Mismatch in Model-based Reinforcement Learningcited by 87
  8. 2020On the role of planning in model-based deep reinforcement learningcited by 57
  9. 2019Information asymmetry in KL-regularized RLcited by 53
  10. 2022Retrieval-Augmented Reinforcement Learningcited by 53
  11. 2022Skill-based Model-based Reinforcement Learningcited by 50
  12. 2020Behavior Priors for Efficient Reinforcement Learningcited by 40
  13. 2017Data-driven Planning via Imitation Learningcited by 36
  14. 2020Model-Predictive Control via Cross-Entropy and Gradient-Based Optimizationcited by 32
  15. 2024Retrieval-Augmented Decision Transformer: External Memory for In-context RLcited by 20
  16. 2026stable-worldmodel: A Platform for Reproducible World Modeling Research and Evaluationcited by 13
  17. 2022Imitating, Fast and Slow: Robust learning from demonstrations via decision-time planningcited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026QQWorld: Quantile-Quantile Matching for World Model Regularizationcited by 5
  2. 2026Aim Short to Reach Far: Your Frozen World Model Can Plan Better Than You Thinkcited by 0
  3. 2026Beyond a single latent space: a dual-latent world model for long-horizon planningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026PRISM: PRior-guided Imagination Sampling in world Models
  2. 2026How Long, Not How Close: A Learned Temporal Metric for Planning in Latent World Models
  3. 2025WorldPrediction: A Benchmark for High-level World Modeling and Long-horizon Procedural Planning
  4. 2026SAGE: Subgoal-Conditioned Action Generation for Latent World Model Planning
  5. 2026Hierarchical Planning with Latent World Models
  6. 2026How Much Planning Is Enough? Reducing Search and Computation in World-Model Planning
  7. 2026Planning in 8 Tokens: A Compact Discrete Tokenizer for Latent World Model
  8. 2026Beyond the Next Step: Variable-Length Latent World Models for Long-Horizon Planning