mlsanityData to 7 Oct 2026

Variational Causal Dynamics: Discovering Modular World Models from Interventions

arXiv 2206.11131 ↗First posted 22 Jun 2022cs.LG stat.MEcited by 16

On the map: World models → World models for language agents

Abstract (arXiv)

Latent world models allow agents to reason about complex environments with high-dimensional observations. However, adapting to new environments and effectively leveraging previous knowledge remain significant challenges. We present variational causal dynamics (VCD), a structured world model that exploits the invariance of causal mechanisms across environments to achieve fast and modular adaptation. By causally factorising a transition model, VCD is able to identify reusable components across different environments. This is achieved by combining causal discovery and variational inference to learn a latent representation and transition model jointly in an unsupervised manner. Specifically, we optimise the evidence lower bound jointly over a representation model and a transition model structured as a causal graphical model. In evaluations on simulated environments with state and image observations, we show that VCD is able to successfully identify causal variables, and to discover consistent causal structures across different environments. Moreover, given a small number of observations in a previously unseen, intervened environment, VCD is able to identify the sparse changes in the dynamics and to adapt efficiently. In doing so, VCD significantly extends the capabilities of the current state-of-the-art in latent world models while also comparing favourably in terms of prediction accuracy.

Key facts. Cited by 16 papers in the map (data through 7 Oct 2026). Placed in World models for language agents, one of 11 directions in World models. Builds on 40 of its 55 references that are in the map. First posted on arXiv 22 Jun 2022.

Builds on

40 of 55 references.

Closest ancestors · 1

Same direction, newest first

  1. 2020Causal World Models by Unsupervised Deconfounding of Physical Dynamics28 Dec

Other references · 39

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2014Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translationcited by 4,016
  3. 2016Categorical Reparameterization with Gumbel-Softmaxcited by 2,708
  4. 2019Invariant Risk Minimizationcited by 1,832
  5. 2019Dream to Control: Learning Behaviors by Latent Imaginationcited by 1,598
  6. 2018Learning Latent Dynamics for Planning from Pixelscited by 1,221
  7. 2018Recurrent World Models Facilitate Policy Evolutioncited by 761
  8. 2015Causal inference using invariant prediction: identification and confidence intervalscited by 533
  9. 2020Planning to Explore via Self-Supervised World Modelscited by 365
  10. 2018Nonlinear ICA Using Auxiliary Variables and Generalized Contrastive Learningcited by 278
  11. 2015Embed to Control: A Locally Linear Latent Dynamics Model for Control from Raw Imagescited by 266
  12. 2018Stochastic Video Generation with a Learned Priorcited by 244
27 more
  1. 2019Recurrent Independent Mechanismscited by 219
  2. 2020Differentiable Causal Discovery from Interventional Datacited by 218
  3. 2019Contrastive Learning of Structured World Modelscited by 206
  4. 2019A Meta-Transfer Objective for Learning to Disentangle Causal Mechanismscited by 158
  5. 2021Disentanglement via Mechanism Sparsity Regularization: A New Principle for Nonlinear ICAcited by 158
  6. 2022Weakly supervised causal representation learningcited by 140
  7. 2022CITRIS: Causal Identifiability from Temporal Intervened Sequencescited by 132
  8. 2019Learning Neural Causal Models from Unknown Interventionscited by 113
  9. 2020CausalWorld: A Robotic Manipulation Benchmark for Causal Structure and Transfer Learningcited by 109
  10. 2021GENESIS-V2: Inferring Unordered Object Representations without Iterative Refinementcited by 109
  11. 2020Invariant Causal Prediction for Block MDPscited by 104
  12. 2021Learning Temporally Causal Latent Processes from General Temporal Datacited by 96
  13. 2020Counterfactual Data Augmentation using Locally Factored Dynamicscited by 93
  14. 2019High Fidelity Video Prediction with Large Stochastic Recurrent Neural Networkscited by 88
  15. 2021Efficient Neural Causal Discovery without Acyclicity Constraintscited by 82
  16. 2019CoPhy: Counterfactual Learning of Physical Dynamicscited by 80
  17. 2020Causal Discovery in Physical Systems from Videoscited by 77
  18. 2021Neural Production Systems: Learning Rule-Governed Visual Dynamicscited by 75
  19. 2021Causal Influence Detection for Improving Efficiency in Reinforcement Learningcited by 70
  20. 2013Causal Discovery from Changescited by 52
  21. 2022Causal Representation Learning for Instantaneous and Temporal Effects in Interactive Systemscited by 50
  22. 2018Learning and Querying Fast Generative Models for Reinforcement Learningcited by 48
  23. 2021AdaRL: What, Where, and How to Adapt in Transfer Reinforcement Learningcited by 46
  24. 2017Robust Locally-Linear Controllable Embeddingcited by 37
  25. 2020Towards causal generative scene models via competition of expertscited by 37
  26. 2021Properties from Mechanisms: An Equivariance Perspective on Identifiable Representation Learningcited by 30
  27. 2021APEX: Unsupervised, Object-Centric Scene Segmentation and Tracking for Robot Manipulationcited by 16

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 1; 2023: 3; 2024: 4; 2025: 1; 2026: 7.

16 papers.

Sort:

By area:

  1. 2023Causal Triplet: An Open Challenge for Intervention-centric Causal Representation Learningcited by 22
  2. 2022Neuro-Symbolic Causal Reasoning Meets Signaling Game for Emergent Semantic Communicationscited by 20
  3. 2024SPARTAN: A Sparse Transformer World Model Attending to What Matterscited by 10
  4. 2024Causal Representation Learning in Temporal Data via Single-Parent Decodingcited by 5
  5. 2026Graph World Models: Concepts, Taxonomy, and Future Directionscited by 5
  6. 2023Out-of-Variable Generalization for Discriminative Modelscited by 4
  7. 2024Compete and Compose: Learning Independent Mechanisms for Modular World Modelscited by 4
  8. 2026From Observation to Insight: Mechanistic World Models and the Quest for Autonomous Discoverycited by 4
  9. 2026Disentangling Dynamical Systems: Causal Representation Learning Meets Local Sparse Attentioncited by 3
  10. 2026Toward Consistent World Models with Multi-Token Prediction and Latent Semantic Enhancementcited by 3
  11. 2023Reliable Beamforming at Terahertz Bands: Are Causal Representations the Way Forward?cited by 2
  12. 2025From Pixels to Factors: Learning Independently Controllable State Variables for Reinforcement Learningcited by 2
4 more
  1. 2026On the Emergence and Test-Time Use of Structural Information in Large Language Modelscited by 1
  2. 2024Disentangled Representations for Causal Cognitioncited by 0
  3. 2026Event-Conditioned Diagnostics of Kinematic, Contact, and Object-Permanence Structure in Passive Object-State World Modelscited by 0
  4. 2026Do-JEPA: From Masking to Intervention in Latent World Modelscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Causal Process Models: Reframing Dynamic Causal Graph Discovery as a Reinforcement Learning Problem
  2. 2025Curious Causality-Seeking Agents Learn Meta Causal World
  3. 2026A Unifying Perspective on Causal World Models: From Observations to Representations to Structure
  4. 2026CausalSteward: An Agentic Divide-Conquer-Combine Copilot for Causal Discovery
  5. 2026Learning Implicit Causal World Models from Multi-Agent Demonstrations
  6. 2025Multi-resolution Score-Based Variational Graphical Diffusion for Causal Disaster System Modeling and Inference
  7. 2026Symmetries and Causality: Causal Effect Identification Beyond IID Data
  8. 2025Learning Local Causal World Models with State Space Models and Attention