mlsanityData to 14 Oct 2026

Learning Causally Disentangled Representations via the Principle of Independent Causal Mechanisms

arXiv 2306.01213 ↗First posted 2 Jun 2023cs.LG stat.MLcited by 14

Not on the map. Why?

Abstract (arXiv)

Learning disentangled causal representations is a challenging problem that has gained significant attention recently due to its implications for extracting meaningful information for downstream tasks. In this work, we define a new notion of causal disentanglement from the perspective of independent causal mechanisms. We propose ICM-VAE, a framework for learning causally disentangled representations supervised by causally related observed labels. We model causal mechanisms using nonlinear learnable flow-based diffeomorphic functions to map noise variables to latent causal variables. Further, to promote the disentanglement of causal factors, we propose a causal disentanglement prior learned from auxiliary labels and the latent causal structure. We theoretically show the identifiability of causal factors and mechanisms up to permutation and elementwise reparameterization. We empirically demonstrate that our framework induces highly disentangled causal factors, improves interventional robustness, and is compatible with counterfactual generation.

Key facts. Cited by 14 papers in the map (data through 14 Oct 2026). Builds on 29 of its 52 references that are in the map. First posted on arXiv 2 Jun 2023.

Builds on

29 of 52 references.

References · 29

  1. 2014Deep Learning Face Attributes in the Wildcited by 3,693
  2. 2019Normalizing Flows for Probabilistic Modeling and Inferencecited by 1,323
  3. 2018Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representationscited by 941
  4. 2018DAGs with NO TEARS: Continuous Optimization for Structure Learningcited by 752
  5. 2019Variational Autoencoders and Nonlinear ICA: A Unifying Frameworkcited by 503
  6. 2021Self-Supervised Learning with Data Augmentations Provably Isolates Content from Stylecited by 312
  7. 2018Nonlinear ICA Using Auxiliary Variables and Generalized Contrastive Learningcited by 282
  8. 2020Weakly-Supervised Disentanglement Without Compromisescited by 239
  9. 2019Gradient-Based Neural DAG Learningcited by 227
  10. 2021D'ya like DAGs? A Survey on Structure Learning and Causal Discoverycited by 211
  11. 2021Disentanglement via Mechanism Sparsity Regularization: A New Principle for Nonlinear ICAcited by 160
  12. 2022Interventional Causal Representation Learningcited by 142
17 more
  1. 2022Weakly supervised causal representation learningcited by 140
  2. 2019DIVA: Domain Invariant Variational Autoencoderscited by 115
  3. 2018Robustly Disentangled Causal Mechanisms: Validating Deep Representations for Interventional Robustnesscited by 108
  4. 2017Learning Independent Causal Mechanismscited by 105
  5. 2020Causal Autoregressive Flowscited by 99
  6. 2021Independent mechanism analysis, a new concept?cited by 97
  7. 2020On Disentangled Representations Learned From Correlated Datacited by 94
  8. 2019Masked Gradient-Based Causal Structure Learningcited by 89
  9. 2019Discrete Flows: Invertible Generative Models of Discrete Datacited by 68
  10. 2020Weakly Supervised Disentangled Generative Causal Representation Learningcited by 68
  11. 2019Causal Discovery with General Non-Linear Relationships Using Non-Linear ICAcited by 58
  12. 2023Causal Component Analysiscited by 56
  13. 2022Causal Representation Learning for Instantaneous and Temporal Effects in Interactive Systemscited by 51
  14. 2023Counterfactual Identifiability of Bijective Causal Modelscited by 43
  15. 2022From Statistical to Causal Learningcited by 40
  16. 2020Capturing Label Characteristics in VAEscited by 30
  17. 2022DCI-ES: An Extended Disentanglement Framework with Connections to Identifiabilitycited by 23

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 3; 2024: 2; 2025: 4; 2026: 5.

14 papers.

Sort:

By area:

  1. 2023Identifiability Guarantees for Causal Disentanglement from Soft Interventionscited by 84
  2. 2023From Identifiable Causal Representations to Controllable Counterfactual Generation: A Survey on Causal Generative Modelingcited by 22
  3. 2024Counterfactual Generative Modeling with Variational Causal Inferencecited by 11
  4. 2025CAUSAL3D: A Comprehensive Benchmark for Causal Learning from Visual Datacited by 7
  5. 2023When Graph Neural Network Meets Causality: Opportunities, Methodologies and An Outlookcited by 4
  6. 2025Improving Generalization Ability of Robotic Imitation Learning by Resolving Causal Confusion in Observationscited by 4
  7. 2025Causal-Adapter: Taming Text-to-Image Diffusion for Faithful Counterfactual Generationcited by 4
  8. 2024Structural Disentanglement of Causal and Correlated Conceptscited by 1
  9. 2026FlexCausal: Flexible Causal Disentanglement via Structural Flow Priors and Manifold-Aware Interventionscited by 1
  10. 2025DISCO: Mitigating Bias in Deep Learning with Conditional Distance Correlationcited by 0
  11. 2026CauCLIP: Bridging the Sim-to-Real Gap in Surgical Video Understanding via Causality-Inspired Vision-Language Modelingcited by 0
  12. 2026A Systematic Framework for Tabular Data Disentanglementcited by 0
2 more
  1. 2026Leveraging Foundation Models for Causal Generative Modelingcited by 0
  2. 2026CouCE: A Unified Causal Framework for Debiased Deep Metric Learningcited by 0