Data to 7 Oct 2026

Weakly supervised causal representation learning

arXiv 2203.16437 ↗First posted 30 Mar 2022stat.ML cs.LGcited by 140

Not on the map. Why?

Abstract (arXiv)

Learning high-level causal representations together with a causal model from unstructured low-level data such as pixels is impossible from observational data alone. We prove under mild assumptions that this representation is however identifiable in a weakly supervised setting. This involves a dataset with paired samples before and after random, unknown interventions, but no further labels. We then introduce implicit latent causal models, variational autoencoders that represent causal variables and causal structure without having to optimize an explicit discrete graph structure. On simple image data, including a novel dataset of simulated robotic manipulation, we demonstrate that such models can reliably identify the causal structure and disentangle causal variables.

Key facts. Cited by 140 papers in the map (data through 7 Oct 2026). Builds on 30 of its 63 references that are in the map. First posted on arXiv 30 Mar 2022.

Builds on

30 of 63 references.

References · 30

2014Adam: A Method for Stochastic Optimizationcited by 36,960
2016SGDR: Stochastic Gradient Descent with Warm Restartscited by 4,993
2016Categorical Reparameterization with Gumbel-Softmaxcited by 2,708
2018Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representationscited by 929
2020Object-Centric Learning with Slot Attentioncited by 865
2018DAGs with NO TEARS: Continuous Optimization for Structure Learningcited by 745
2019Variational Autoencoders and Nonlinear ICA: A Unifying Frameworkcited by 494
2015Importance Weighted Autoencoderscited by 454
2021Self-Supervised Learning with Data Augmentations Provably Isolates Content from Stylecited by 308
2021Conditional Object-Centric Learning from Videocited by 247
2020Weakly-Supervised Disentanglement Without Compromisescited by 237
2020Differentiable Causal Discovery from Interventional Datacited by 218

18 more

2021Towards Causal Representation Learningcited by 207
2021Disentanglement via Mechanism Sparsity Regularization: A New Principle for Nonlinear ICAcited by 158
2016Foundations of Structural Causal Models with Cycles and Latent Variablescited by 134
2022CITRIS: Causal Identifiability from Temporal Intervened Sequencescited by 132
2020Flows for simultaneous manifold learning and density estimationcited by 117
2017Causal Consistency of Structural Equation Modelscited by 114
2018Abstracting Causal Modelscited by 104
2021Independent mechanism analysis, a new concept?cited by 96
2021Learning Temporally Causal Latent Processes from General Temporal Datacited by 96
2019Weakly Supervised Disentanglement with Guaranteescited by 82
2021Efficient Neural Causal Discovery without Acyclicity Constraintscited by 82
2021Disentangling Identifiable Features from Noisy Data with Structured Nonlinear ICAcited by 64
2014Visual Causal Feature Learningcited by 59
2019A synthetic approach to Markov kernels, conditional independence and theorems on sufficient statisticscited by 55
2021Learning latent causal graphs via mixture oraclescited by 55
2022Differentiable DAG Samplingcited by 46
2021Nonlinear Invariant Risk Minimization: A Causal Approachcited by 42
2020TriFinger: An Open-Source Robot for Learning Dexteritycited by 31

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 11; 2023: 39; 2024: 32; 2025: 27; 2026: 31.

140 papers.

Sort:

By area:

2022Interventional Causal Representation Learningcited by 140
2023Nonparametric Identifiability of Causal Representations from Unknown Interventionscited by 105
2023Learning Linear Causal Representations from Interventions under General Nonlinear Mixingcited by 92
2022Linear Causal Disentanglement via Interventionscited by 84
2023Identifiability Guarantees for Causal Disentanglement from Soft Interventionscited by 84
2023Multi-View Causal Representation Learning with Partial Observabilitycited by 72
2022Identifiability of deep generative models without auxiliary informationcited by 70
2024Causal Representation Learning from Multiple Distributions: A General Settingcited by 67
2022Weakly Supervised Representation Learning with Sparse Perturbationscited by 64
2023Identifiability Results for Multimodal Contrastive Learningcited by 58
2023Causal Component Analysiscited by 56
2023Additive Decoders for Latent Variables Identification and Cartesian-Product Extrapolationcited by 56

38 more

2022Causal Representation Learning for Instantaneous and Temporal Effects in Interactive Systemscited by 50
2023Score-based Causal Representation Learning with Interventionscited by 49
2022Synergies between Disentanglement and Sparsity: Generalization and Identifiability in Multi-Task Learningcited by 44
2023BISCUIT: Causal Representation Learning from Binary Interactionscited by 38
2023Learning nonparametric latent causal graphs with unknown interventionscited by 34
2023General Identifiability and Achievability for Causal Representation Learningcited by 34
2024A Sparsity Principle for Partially Observable Causal Representation Learningcited by 34
2022Indeterminacy in Generative Models: Characterization and Strong Identifiabilitycited by 33
2022Identifying Weight-Variant Latent Causal Modelscited by 33
2023Leveraging sparse and shared feature activations for disentangled representation learningcited by 32
2023Causal Representation Learning Made Identifiable by Grouping of Observational Variablescited by 30
2024Unifying Causal Representation Learning with the Invariance Principlecited by 30
2024Learning Interpretable Concepts: Unifying Causal Representation Learning and Foundation Modelscited by 29
2025Thought Communication in Multiagent Collaborationcited by 28
2023Multi-Domain Causal Representation Learning via Weak Distributional Invariancescited by 23
2023Identifiable Latent Polynomial Causal Models Through the Lens of Changecited by 23
2023Causal Triplet: An Open Challenge for Intervention-centric Causal Representation Learningcited by 22
2024Marrying Causal Representation Learning with Dynamical Systems for Sciencecited by 22
2023Identifying Representations for Intervention Extrapolationcited by 21
2023From Identifiable Causal Representations to Controllable Counterfactual Generation: A Survey on Causal Generative Modelingcited by 21
2023Object-centric architectures enable efficient causal representation learningcited by 21
2024Counterfactual Image Editingcited by 20
2024Interaction Asymmetry: A General Principle for Learning Composable Abstractionscited by 19
2023Identifying Linearly-Mixed Causal Representations from Multi-Node Interventionscited by 18
2024Causal Diffusion Autoencoders: Toward Counterfactual Generation via Diffusion Probabilistic Modelscited by 17
2024The Landscape of Causal Discovery Data: Grounding Causal Discovery in Real-World Applicationscited by 17
2022Variational Causal Dynamics: Discovering Modular World Models from Interventionscited by 16
2023Self-Supervised Disentanglement by Leveraging Structure in Data Augmentationscited by 16
2023Causal models in string diagramscited by 14
2023Learning Causally Disentangled Representations via the Principle of Independent Causal Mechanismscited by 14
2024Linear Causal Representation Learning from Unknown Multi-node Interventionscited by 14
2023Towards a Unified Framework of Contrastive Learning for Disentangled Representationscited by 13
2024A Complexity-Based Theory of Compositionalitycited by 13
2025I Predict Therefore I Am: Is Next Token Prediction Enough to Learn Human-Interpretable Concepts from Data?cited by 13
2024Smoke and Mirrors in Causal Downstream Taskscited by 12
2024Identifiable Causal Representation Learning: Unsupervised, Multi-View, and Multi-Environmentcited by 12
2025Towards Interpretable Deep Generative Models via Causal Representation Learningcited by 12
2024Counterfactual Generative Modeling with Variational Causal Inferencecited by 10

50 most cited of 140 All 140 (JSON)