mlsanityData to 7 Oct 2026

Reinformed Dreamer: An Asymmetric World Model Efficiently Trained through Latent Guidance

arXiv 2607.26040 ↗First posted 28 Jul 2026cs.LG stat.MLcited by 0

On the map: World models → Latent dynamics models and model-based RL

Abstract (arXiv)

Much like humans benefit from guidance while learning, reinforcement learning algorithms may benefit from additional supervision beyond rewards. Leveraging additional information during training to learn better representations and behaviors has been the focus of asymmetric reinforcement learning. This learning paradigm has proven effective under partial observability when additional state information is available, but also under full observability when more refined state information is available. Focusing on model-based reinforcement learning, we study the effect of asymmetric learning on observation representations and on privileged information representations. First, we identify a limitation in the privileged information representations learned by an asymmetric model-based algorithm known as the Informed Dreamer. Then, we propose a novel asymmetric representation learning objective using latent guidance, resulting in a new algorithm called the Reinformed Dreamer. Experiments across several benchmarks show a more consistent improvement over Dreamer than previous asymmetric approaches.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Latent dynamics models and model-based RL, one of 11 directions in World models. Builds on 47 of its 68 references that are in the map. First posted on arXiv 28 Jul 2026.

Builds on

47 of 68 references.

Closest ancestors · 5

Same direction, newest first

  1. 2024Mastering Memory Tasks with World Models7 Mar
  2. 2023Informed POMDP: Leveraging Additional Information in Model-Based RL20 Jun
  3. 2020Mastering Atari with Discrete World Models5 Oct
  4. 2019Dream to Control: Learning Behaviors by Latent Imagination3 Dec
  5. 2018Recurrent World Models Facilitate Policy Evolution4 Sep

Other references · 42

  1. 2017Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environmentscited by 1,519
  2. 2018Learning Latent Dynamics for Planning from Pixelscited by 1,221
  3. 2018DeepMind Control Suitecited by 935
  4. 2017Counterfactual Multi-Agent Policy Gradientscited by 733
  5. 2018QMIX: Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learningcited by 697
  6. 2015Deep Recurrent Q-Learning for Partially Observable MDPscited by 457
  7. 2019QTRAN: Learning to Factorize with Transformation for Cooperative Multi-Agent Reinforcement Learningcited by 372
  8. 2020QPLEX: Duplex Dueling Multi-Agent Q-Learningcited by 267
  9. 2019Stochastic Latent Actor-Critic: Deep Reinforcement Learning with a Latent Variable Modelcited by 228
  10. 2017Asymmetric Actor Critic for Image-Based Robot Learningcited by 202
  11. 2020Weighted QMIX: Expanding Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learningcited by 197
  12. 2019Lagging Inference Networks and Posterior Collapse in Variational Autoencoderscited by 124
30 more
  1. 2018Deep Variational Reinforcement Learning for POMDPscited by 122
  2. 2020Bootstrap Latent-Predictive Representations for Multitask Reinforcement Learningcited by 101
  3. 2015Memory-based control with recurrent neural networkscited by 77
  4. 2020Approximate information state for approximate planning and reinforcement learning in partially observed systemscited by 67
  5. 2019Shaping Belief States with Generative Environment Models for RLcited by 65
  6. 2023POPGym: Benchmarking Partially Observable Reinforcement Learningcited by 59
  7. 2024Bridging State and History Representations: Understanding Self-Predictive RLcited by 52
  8. 2018On Improving Deep Reinforcement Learning for POMDPscited by 49
  9. 2018Learning and Querying Fast Generative Models for Reinforcement Learningcited by 48
  10. 2019Variational Recurrent Models for Solving Partially Observable Control Taskscited by 46
  11. 2024An Introduction to Centralized Training for Decentralized Execution in Cooperative Multi-Agent Reinforcement Learningcited by 46
  12. 2018Neural Predictive Belief Representationscited by 39
  13. 2017Data-driven Planning via Imitation Learningcited by 36
  14. 2020R-MADDPG for Partially Observable Environments and Limited Communicationcited by 36
  15. 2021Unbiased Asymmetric Reinforcement Learning under Partial Observabilitycited by 24
  16. 2020Robust Asymmetric Learning in POMDPscited by 23
  17. 2024Privileged Sensing Scaffolds Reinforcement Learningcited by 18
  18. 2022Rethinking Individual Global Max in Cooperative Multi-Agent Reinforcement Learningcited by 13
  19. 2022On Uncertainty in Deep State Space Models for Model-Based Reinforcement Learningcited by 13
  20. 2024Provable Partially Observable Reinforcement Learning with Privileged Informationcited by 13
  21. 2022Finite-Time Analysis of Natural Actor-Critic for POMDPscited by 12
  22. 2024A Super-human Vision-based Reinforcement Learning Agent for Autonomous Racing in Gran Turismocited by 12
  23. 2022Recurrent networks, hidden states and beliefs in partially observable environmentscited by 11
  24. 2022A Deeper Understanding of State-Based Critics in Multi-Agent Reinforcement Learningcited by 9
  25. 2023The Wasserstein Believer: Learning Belief Updates for Partially Observable Environments through Reliable Latent Space Modelscited by 9
  26. 2024Recurrent Natural Policy Gradient for POMDPscited by 8
  27. 2020QVMix and QVMix-Max: Extending the Deep Quality-Value Family of Algorithms to Cooperative Multi-Agent Reinforcement Learningcited by 5
  28. 2025A Theoretical Justification for Asymmetric Actor-Critic Algorithmscited by 5
  29. 2025Informed Asymmetric Actor-Critic: Leveraging Privileged Signals Beyond Full-State Accesscited by 2
  30. 2025Real-World Reinforcement Learning of Active Perception Behaviorscited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2021DreamerPro: Reconstruction-Free Model-Based Reinforcement Learning with Prototypical Representations
  2. 2025PIGDreamer: Privileged Information Guided World Models for Safe Partially Observable Reinforcement Learning
  3. 2026Dreamer-CDP: Improving Reconstruction-free World Models Via Continuous Deterministic Representation Prediction
  4. 2025Zero-shot World Models via Search in Memory
  5. 2020Dreaming: Model-based Reinforcement Learning by Latent Imagination without Reconstruction
  6. 2026R2-Dreamer: Redundancy-Reduced World Models without Decoders or Augmentation
  7. 2019Learning to Predict Without Looking Ahead: World Models Without Forward Prediction
  8. 2024Learning Latent Dynamic Robust Representations for World Models