mlsanityData to 7 Oct 2026

Identifiability Guarantees for Causal Disentanglement from Soft Interventions

arXiv 2307.06250 ↗First posted 12 Jul 2023stat.ML cs.LG math.STcited by 84

Not on the map. Why?

Abstract (arXiv)

Causal disentanglement aims to uncover a representation of data using latent variables that are interrelated through a causal model. Such a representation is identifiable if the latent model that explains the data is unique. In this paper, we focus on the scenario where unpaired observational and interventional data are available, with each intervention changing the mechanism of a latent variable. When the causal variables are fully observed, statistically consistent algorithms have been developed to identify the causal model under faithfulness assumptions. We here show that identifiability can still be achieved with unobserved causal variables, given a generalized notion of faithfulness. Our results guarantee that we can recover the latent causal model up to an equivalence class and predict the effect of unseen combinations of interventions, in the limit of infinite data. We implement our causal disentanglement framework by developing an autoencoding variational Bayes algorithm and apply it to the problem of predicting combinatorial perturbation effects in genomics.

Key facts. Cited by 84 papers in the map (data through 7 Oct 2026). Builds on 38 of its 73 references that are in the map. First posted on arXiv 12 Jul 2023.

Builds on

38 of 73 references.

References · 38

  1. 2018UMAP: Uniform Manifold Approximation and Projection for Dimension Reductioncited by 2,841
  2. 2019Invariant Risk Minimizationcited by 1,832
  3. 2018DAGs with NO TEARS: Continuous Optimization for Structure Learningcited by 745
  4. 2019Variational Autoencoders and Nonlinear ICA: A Unifying Frameworkcited by 494
  5. 2018Nonlinear ICA Using Auxiliary Variables and Generalized Contrastive Learningcited by 278
  6. 2019Gradient-Based Neural DAG Learningcited by 227
  7. 2021Contrastive Learning Inverts the Data Generating Processcited by 220
  8. 2020Deep Structural Causal Models for Tractable Counterfactual Inferencecited by 210
  9. 2021D'ya like DAGs? A Survey on Structure Learning and Causal Discoverycited by 209
  10. 2022Weakly supervised causal representation learningcited by 140
  11. 2022Interventional Causal Representation Learningcited by 140
  12. 2022CITRIS: Causal Identifiability from Temporal Intervened Sequencescited by 132
26 more
  1. 2022Causal Machine Learning: A Survey and Open Problemscited by 107
  2. 2021DiBS: Differentiable Bayesian Structure Learningcited by 105
  3. 2023Nonparametric Identifiability of Causal Representations from Unknown Interventionscited by 105
  4. 2023Learning Linear Causal Representations from Interventions under General Nonlinear Mixingcited by 92
  5. 2017Causal Structure Learningcited by 84
  6. 2022Linear Causal Disentanglement via Interventionscited by 84
  7. 2017Permutation-based Causal Inference Algorithms with Interventionscited by 78
  8. 2018Characterizing and Learning Equivalence Classes of Causal DAGs under Interventionscited by 76
  9. 2022Weakly Supervised Representation Learning with Sparse Perturbationscited by 64
  10. 2019Sliced Wasserstein Generative Modelscited by 61
  11. 2023Causal Component Analysiscited by 56
  12. 2021Learning latent causal graphs via mixture oraclescited by 55
  13. 2013Causal Discovery from Changescited by 52
  14. 2023Score-based Causal Representation Learning with Interventionscited by 49
  15. 2021Unsuitability of NOTEARS for Causal Graph Discoverycited by 46
  16. 2022Causal Structure Learning: a Combinatorial Perspectivecited by 46
  17. 2015backShift: Learning causal cyclic graphs from unknown shift interventionscited by 37
  18. 2023Learning nonparametric latent causal graphs with unknown interventionscited by 34
  19. 2021Properties from Mechanisms: An Equivariance Perspective on Identifiable Representation Learningcited by 30
  20. 2023Identification of Nonlinear Latent Hierarchical Modelscited by 27
  21. 2021Matching a Desired Causal State via Shift Interventionscited by 16
  22. 2019Learning Embeddings into Entropic Wasserstein Spacescited by 15
  23. 2023Learning Causally Disentangled Representations via the Principle of Independent Causal Mechanismscited by 14
  24. 2015Anchored Discrete Factor Analysiscited by 9
  25. 2022The interventional Bayesian Gaussian equivalent score for Bayesian causal inference with unknown soft interventionscited by 4
  26. 2023Neuro-Causal Factor Analysiscited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 8; 2024: 29; 2025: 26; 2026: 21.

84 papers.

Sort:

By area:

  1. 2023Multi-View Causal Representation Learning with Partial Observabilitycited by 72
  2. 2024Causal Representation Learning from Multiple Distributions: A General Settingcited by 67
  3. 2023General Identifiability and Achievability for Causal Representation Learningcited by 34
  4. 2024A Sparsity Principle for Partially Observable Causal Representation Learningcited by 34
  5. 2024Unifying Causal Representation Learning with the Invariance Principlecited by 30
  6. 2024Learning Interpretable Concepts: Unifying Causal Representation Learning and Foundation Modelscited by 29
  7. 2023Multi-Domain Causal Representation Learning via Weak Distributional Invariancescited by 23
  8. 2024Marrying Causal Representation Learning with Dynamical Systems for Sciencecited by 22
  9. 2023Identifying Representations for Intervention Extrapolationcited by 21
  10. 2024Counterfactual Image Editingcited by 20
  11. 2023Identifying Linearly-Mixed Causal Representations from Multi-Node Interventionscited by 18
  12. 2024The Landscape of Causal Discovery Data: Grounding Causal Discovery in Real-World Applicationscited by 17
38 more
  1. 2023Learning Causal Representations from General Environments: Identifiability and Intrinsic Ambiguitycited by 14
  2. 2024Linear Causal Representation Learning from Unknown Multi-node Interventionscited by 14
  3. 2024All or None: Identifiable Linear Properties of Next-token Predictors in Language Modelingcited by 14
  4. 2025SmartCLIP: Modular Vision-language Alignment with Identification Guaranteescited by 14
  5. 2024Smoke and Mirrors in Causal Downstream Taskscited by 12
  6. 2024Identifiable Causal Representation Learning: Unsupervised, Multi-View, and Multi-Environmentcited by 12
  7. 2024Causal Representation Learning from Multimodal Biomedical Observationscited by 12
  8. 2025Towards Interpretable Deep Generative Models via Causal Representation Learningcited by 12
  9. 2024Season combinatorial intervention predictions with Salt & Pepercited by 11
  10. 2023An Interventional Perspective on Identifiability in Gaussian LTI Systems with Independent Component Analysiscited by 9
  11. 2024Automated Discovery of Pairwise Interactions from Unstructured Datacited by 8
  12. 2024Language Agents Meet Causality -- Bridging LLMs and Causal World Modelscited by 8
  13. 2024Identifiability Guarantees for Causal Disentanglement from Purely Observational Datacited by 7
  14. 2025CausalVLBench: Benchmarking Visual Causal Reasoning in Large Vision-Language Modelscited by 7
  15. 2025What is causal about causal models and representations?cited by 6
  16. 2025Debiasing Reward Models by Representation Learning with Guaranteescited by 6
  17. 2023Invariance & Causal Representation Learning: Prospects and Limitationscited by 5
  18. 2024Towards Identifiable Latent Additive Noise Modelscited by 5
  19. 2024Towards the Causal Complete Cause of Multi-Modal Representation Learningcited by 5
  20. 2025Sanity Checking Causal Representation Learning on a Simple Real-World Systemcited by 5
  21. 2025From Isolation to Entanglement: When Do Interpretability Methods Identify and Disentangle Known Concepts?cited by 5
  22. 2024Toward the Identifiability of Comparative Deep Generative Modelscited by 4
  23. 2024Amortized Inference of Causal Models via Conditional Fixed-Point Iterationscited by 4
  24. 2024Learning Identifiable Factorized Causal Representations of Cellular Responsescited by 4
  25. 2025The Third Pillar of Causal Analysis? A Measurement Perspective on Causal Representationscited by 4
  26. 2025When Does Closeness in Distribution Imply Representational Similarity? An Identifiability Perspectivecited by 4
  27. 2025Discovering Hierarchical Latent Capabilities of Language Models via Causal Representation Learningcited by 4
  28. 2025Exploratory Causal Inference in SAEncecited by 4
  29. 2026Distributional Equivalence in Linear Non-Gaussian Latent-Variable Cyclic Causal Models: Characterization and Learningcited by 4
  30. 2026Causal Representation Learning from General Environments under Nonparametric Mixingcited by 4
  31. 2025Mechanistic Independence: A Principle for Identifiable Disentangled Representationscited by 3
  32. 2025Causal Structure and Representation Learning with Biomedical Applicationscited by 3
  33. 2026What Makes a Representation Good for Single-Cell Perturbation Prediction?cited by 3
  34. 2024Structured Learning of Compositional Sequential Interventionscited by 2
  35. 2024Cross-validating causal discovery via Leave-One-Variable-Outcited by 2
  36. 2025From Invariant Representations to Invariant Data: Provable Robustness to Spurious Correlations via Noisy Counterfactual Matchingcited by 2
  37. 2025On the Identifiability of Latent Action Policiescited by 2
  38. 2025From Pixels to Factors: Learning Independently Controllable State Variables for Reinforcement Learningcited by 2

50 most cited of 84 All 84 (JSON)