mlsanityData to 14 Oct 2026

Looped World Models

arXiv 2606.18208 ↗First posted 16 Jun 2026cs.LG cs.AI cs.CLcited by 3

On the map: World models → Latent predictive world models

Abstract (arXiv)

Current world models face a fundamental tension: faithful long-horizon simulation demands deep computation, but deeper models are expensive to deploy and prone to compounding errors. We resolve this by introducing Looped World Models (LoopWM), which are the first looped architectures for world modelling. Our method iteratively refines latent environment states through a parameter-shared transformer block. This yield up to 100x parameter efficiency over conventional approaches with adaptive computation that automatically scales depth to match the complexity of each prediction step. Orthogonal to scaling model size and training data, LoopWM establishes iterative latent depth as a new scaling axis for world simulation, which might significantly push the community forward.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Placed in Latent predictive world models, one of 11 directions in World models. Builds on 46 of its 52 references that are in the map. First posted on arXiv 16 Jun 2026.

Builds on

46 of 52 references.

Closest ancestors · 1

Same direction, newest first

  1. 2018Learning Latent Dynamics for Planning from Pixels12 Nov

Other references · 45

  1. 2018Neural Ordinary Differential Equationscited by 3,779
  2. 2019ALBERT: A Lite BERT for Self-supervised Learning of Language Representationscited by 2,790
  3. 2019Dream to Control: Learning Behaviors by Latent Imaginationcited by 1,653
  4. 2019Mastering Atari, Go, Chess and Shogi by Planning with a Learned Modelcited by 1,353
  5. 2020Mastering Atari with Discrete World Modelscited by 1,007
  6. 2018Recurrent World Models Facilitate Policy Evolutioncited by 785
  7. 2024Genie: Generative Interactive Environmentscited by 747
  8. 2018Universal Transformerscited by 647
  9. 2019Deep Equilibrium Modelscited by 615
  10. 2017BranchyNet: Fast Inference via Early Exiting from Deep Neural Networkscited by 596
  11. 2016Adaptive Computation Time for Recurrent Neural Networkscited by 506
  12. 2019Model-Based Reinforcement Learning for Ataricited by 502
33 more
  1. 2025Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approachcited by 415
  2. 2022Transformers are Sample-Efficient World Modelscited by 328
  3. 2022ScienceWorld: Is your Agent Smarter than a 5th Grader?cited by 327
  4. 2018TextWorld: A Learning Environment for Text-based Gamescited by 310
  5. 2024Diffusion for World Modeling: Visual Details Matter in Ataricited by 300
  6. 2025Reasoning with Latent Thoughts: On the Power of Looped Transformerscited by 240
  7. 2023Looped Transformers as Programmable Computerscited by 237
  8. 2025Scaling Latent Reasoning via Looped Language Modelscited by 165
  9. 2017Adaptive Neural Networks for Efficient Inferencecited by 163
  10. 2023Looped Transformers are Better at Learning Learning Algorithmscited by 149
  11. 2022TransDreamer: Reinforcement Learning with Transformer World Modelscited by 143
  12. 2025Mixture-of-Recursions: Learning Dynamic Recursive Depths for Adaptive Token-Level Computationcited by 133
  13. 2024Looped Transformers for Length Generalizationcited by 99
  14. 2024World Models for Autonomous Driving: An Initial Surveycited by 80
  15. 2025A Comprehensive Survey on World Models for Embodied AIcited by 68
  16. 2022A Survey on Model-based Reinforcement Learningcited by 66
  17. 2026LoopFormer: Elastic-Depth Looped Transformers for Latent Reasoning via Shortcut Modulationcited by 61
  18. 2025A Survey of World Models for Autonomous Drivingcited by 60
  19. 2024MoEUT: Mixture-of-Experts Universal Transformerscited by 59
  20. 2026Parcae: Scaling Laws For Stable Looped Language Modelscited by 55
  21. 2024Sora as a World Model? A Complete Survey on Text-to-Video Generationcited by 53
  22. 2016Self-Correcting Models for Model-Based Reinforcement Learningcited by 43
  23. 2019Learning to Combat Compounding-Error in Model-Based Reinforcement Learningcited by 39
  24. 2024Efficient World Models with Context-Aware Tokenizationcited by 33
  25. 2025Encode, Think, Decode: Scaling test-time reasoning with recursive latent thoughtscited by 24
  26. 2025PIN-WM: Learning Physics-INformed World Models for Non-Prehensile Manipulationcited by 22
  27. 2025Two-Scale Latent Dynamics for Recurrent-Depth Transformerscited by 13
  28. 2026Hyperloop Transformerscited by 12
  29. 2025ProPhy: Progressive Physical Alignment for Dynamic World Simulationcited by 11
  30. 2025A Survey of Early Exit Deep Neural Networks in NLPcited by 10
  31. 2025Accurate and Efficient World Modeling with Masked Latent Transformerscited by 7
  32. 2026A Mechanistic View on Video Generation as World Models: State and Dynamicscited by 6
  33. 2026Next Embedding Prediction Makes World Models Strongercited by 4

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026Situation Perception: A Necessary Primitive to Artificial Superintelligencecited by 0
  2. 2026Looped Actor: Depth-Recurrent Reasoning Models for Reinforcement Learningcited by 0
  3. 2026LoopMoEVR: Loop-Based Degradation-Aware Mixture-of-Experts for Unified UHD Video Restorationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026stable-worldmodel-v1: Reproducible World Modeling Research and Evaluation
  2. 2026stable-worldmodel: A Platform for Reproducible World Modeling Research and Evaluation
  3. 2026An Empirical Study of World Model Quantization
  4. 2024D2-World: An Efficient World Model through Decoupled Dynamic Flow
  5. 2024One-shot World Models Using a Transformer Trained on a Synthetic Prior
  6. 2025World-in-World: World Models in a Closed-Loop World
  7. 2026Beyond Myopic World Models: Long-Horizon End-to-End Training for Direct Future Prediction
  8. 2025SmallWorlds: Assessing Dynamics Understanding of World Models in Isolated Environments