mlsanityData to 14 Oct 2026

Simple, Good, Fast: Self-Supervised World Models Free of Baggage

arXiv 2506.02612 ↗First posted 3 Jun 2025cs.LG cs.AI stat.MLcited by 3

On the map: World models → Video-generation world models and simulators

Abstract (arXiv)

What are the essential components of world models? How far do we get with world models that are not employing RNNs, transformers, discrete representations, and image reconstructions? This paper introduces SGF, a Simple, Good, and Fast world model that uses self-supervised representation learning, captures short-time dependencies through frame and action stacking, and enhances robustness against model errors through data augmentation. We extensively discuss SGF's connections to established world models, evaluate the building blocks in ablation studies, and demonstrate good performance through quantitative comparisons on the Atari 100k benchmark.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Placed in Video-generation world models and simulators, one of 11 directions in World models. Builds on 50 of its 72 references that are in the map. First posted on arXiv 3 Jun 2025.

Builds on

50 of 72 references.

References · 50

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2017Proximal Policy Optimization Algorithmscited by 15,215
  3. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 12,049
  4. 2018Representation Learning with Contrastive Predictive Codingcited by 7,777
  5. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,513
  6. 2015Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shiftcited by 7,005
  7. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,968
  8. 2016Layer Normalizationcited by 5,669
  9. 2020Bootstrap your own latent: A new approach to self-supervised Learningcited by 4,750
  10. 2020Exploring Simple Siamese Representation Learningcited by 2,866
  11. 2020Unsupervised Learning of Visual Features by Contrasting Cluster Assignmentscited by 2,857
  12. 2016Asynchronous Methods for Deep Reinforcement Learningcited by 2,258
38 more
  1. 2013Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computationcited by 2,079
  2. 2015High-Dimensional Continuous Control Using Generalized Advantage Estimationcited by 2,026
  3. 2021Barlow Twins: Self-Supervised Learning via Redundancy Reductioncited by 1,838
  4. 2019Dream to Control: Learning Behaviors by Latent Imaginationcited by 1,653
  5. 2019Mastering Atari, Go, Chess and Shogi by Planning with a Learned Modelcited by 1,353
  6. 2018Learning Latent Dynamics for Planning from Pixelscited by 1,264
  7. 2023Mastering Diverse Domains through World Modelscited by 1,253
  8. 2021VICReg: Variance-Invariance-Covariance Regularization for Self-Supervised Learningcited by 1,057
  9. 2020Mastering Atari with Discrete World Modelscited by 1,007
  10. 2021Deep Reinforcement Learning at the Edge of the Statistical Precipicecited by 807
  11. 2017Rainbow: Combining Improvements in Deep Reinforcement Learningcited by 786
  12. 2018Recurrent World Models Facilitate Policy Evolutioncited by 785
  13. 2018Deep Reinforcement Learning in a Handful of Trials using Probabilistic Dynamics Modelscited by 774
  14. 2020CURL: Contrastive Unsupervised Representations for Reinforcement Learningcited by 760
  15. 2019Data-Efficient Image Recognition with Contrastive Predictive Codingcited by 673
  16. 2020Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixelscited by 585
  17. 2019Model-Based Reinforcement Learning for Ataricited by 502
  18. 2020Reinforcement Learning with Augmented Datacited by 468
  19. 2020Learning Invariant Representations for Reinforcement Learning without Reconstructioncited by 436
  20. 2022Temporal Difference Learning for Model Predictive Controlcited by 397
  21. 2022Transformers are Sample-Efficient World Modelscited by 328
  22. 2020Data-Efficient Reinforcement Learning with Self-Predictive Representationscited by 318
  23. 2020Agent57: Outperforming the Atari Human Benchmarkcited by 303
  24. 2015Embed to Control: A Locally Linear Latent Dynamics Model for Control from Raw Imagescited by 274
  25. 2017Revisiting the Arcade Learning Environment: Evaluation Protocols and Open Problems for General Agentscited by 269
  26. 2021Mastering Atari Games with Limited Datacited by 237
  27. 2019Stochastic Latent Actor-Critic: Deep Reinforcement Learning with a Latent Variable Modelcited by 231
  28. 2019DeepMDP: Learning Continuous Latent Space Models for Representation Learningcited by 207
  29. 2016Value Iteration Networkscited by 160
  30. 2023Bigger, Better, Faster: Human-level Atari with human-level efficiencycited by 147
  31. 2023Transformer-based World Models Are Happy With 100k Interactionscited by 139
  32. 2019When to use parametric models in reinforcement learning?cited by 118
  33. 2017Value Prediction Networkcited by 117
  34. 2022On the duality between contrastive and non-contrastive self-supervised learningcited by 108
  35. 2021Pretraining Representations for Data-Efficient Reinforcement Learningcited by 105
  36. 2016The Predictron: End-To-End Learning and Planningcited by 93
  37. 2017Robust Locally-Linear Controllable Embeddingcited by 38
  38. 2022Human-level Atari 200x fastercited by 28

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 2; 2026: 1.

3 papers.

  1. 2025Edge General Intelligence Through World Models and Agentic AI: Fundamentals, Solutions, and Challengescited by 3
  2. 2025One Period to Rule Them All: Identifying Critical Learning Periods in Deep Networkscited by 1
  3. 2026Understanding and Stabilizing Deep Q-Learning via Controlled Bootstrapping and Regulated Value Dynamicscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Latent Video Prediction Learns Better World Models
  2. 2024Learning and Leveraging World Models in Visual Representation Learning
  3. 2026SANA-WM: Efficient Minute-Scale World Modeling with Hybrid Linear Diffusion Transformer
  4. 2025Can World Models Benefit VLMs for World Dynamics?
  5. 2026Latent Particle World Models: Self-supervised Object-centric Stochastic Dynamics Modeling
  6. 2025SmallWorlds: Assessing Dynamics Understanding of World Models in Isolated Environments
  7. 2026SolarWM: Open Data and Scalable Training for Long-Horizon Video World Models
  8. 2025WorldModelBench: Judging Video Generation Models As World Models