mlsanityData to 7 Oct 2026

Towards the Reusability and Compositionality of Causal Representations

arXiv 2403.09830 ↗First posted 14 Mar 2024cs.LG cs.AIcited by 5

Not on the map. Why?

Abstract (arXiv)

Causal Representation Learning (CRL) aims at identifying high-level causal factors and their relationships from high-dimensional observations, e.g., images. While most CRL works focus on learning causal representations in a single environment, in this work we instead propose a first step towards learning causal representations from temporal sequences of images that can be adapted in a new environment, or composed across multiple related environments. In particular, we introduce DECAF, a framework that detects which causal factors can be reused and which need to be adapted from previously learned causal representations. Our approach is based on the availability of intervention targets, that indicate which variables are perturbed at each time step. Experiments on three benchmark datasets show that integrating our framework with four state-of-the-art CRL approaches leads to accurate representations in a new environment with only a few samples.

Key facts. Cited by 5 papers in the map (data through 7 Oct 2026). Builds on 31 of its 49 references that are in the map. First posted on arXiv 14 Mar 2024.

Builds on

31 of 49 references.

References · 31

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2016Gaussian Error Linear Units (GELUs)cited by 3,431
  3. 2015Variational Inference with Normalizing Flowscited by 2,161
  4. 2018Glow: Generative Flow with Invertible 1x1 Convolutionscited by 1,539
  5. 2018Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representationscited by 929
  6. 2016Improving Variational Inference with Inverse Autoregressive Flowcited by 636
  7. 2019Variational Autoencoders and Nonlinear ICA: A Unifying Frameworkcited by 494
  8. 2021Self-Supervised Learning with Data Augmentations Provably Isolates Content from Stylecited by 308
  9. 2015MADE: Masked Autoencoder for Distribution Estimationcited by 306
  10. 2018Nonlinear ICA Using Auxiliary Variables and Generalized Contrastive Learningcited by 278
  11. 2016Unsupervised Feature Extraction by Time-Contrastive Learning and Nonlinear ICAcited by 270
  12. 2020Weakly-Supervised Disentanglement Without Compromisescited by 237
19 more
  1. 2019Recurrent Independent Mechanismscited by 219
  2. 2021Disentanglement via Mechanism Sparsity Regularization: A New Principle for Nonlinear ICAcited by 158
  3. 2019On the Fairness of Disentangled Representationscited by 144
  4. 2022Weakly supervised causal representation learningcited by 140
  5. 2022CITRIS: Causal Identifiability from Temporal Intervened Sequencescited by 132
  6. 2017Learning Independent Causal Mechanismscited by 104
  7. 2019Are Disentangled Representations Helpful for Abstract Visual Reasoning?cited by 100
  8. 2021Independent mechanism analysis, a new concept?cited by 96
  9. 2021Learning Temporally Causal Latent Processes from General Temporal Datacited by 96
  10. 2020On Disentangled Representations Learned From Correlated Datacited by 94
  11. 2022On the Identifiability of Nonlinear ICA: Sparsity and Beyondcited by 84
  12. 2021Generalization and Robustness Implications in Object-Centric Learningcited by 81
  13. 2015Learning Causal Graphs with Small Interventionscited by 63
  14. 2020On the Transfer of Disentangled Representations in Realistic Settingscited by 63
  15. 2022Temporally Disentangled Representation Learningcited by 59
  16. 2018Counterfactuals uncover the modular structure of deep generative modelscited by 51
  17. 2022Causal Representation Learning for Instantaneous and Temporal Effects in Interactive Systemscited by 50
  18. 2023BISCUIT: Causal Representation Learning from Binary Interactionscited by 38
  19. 2022Disentanglement of Correlated Factors via Hausdorff Factorized Supportcited by 37

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 2; 2025: 2; 2026: 1.

5 papers.

  1. 2025Position: We Need An Algorithmic Understanding of Generative AIcited by 5
  2. 2024Systems with Switching Causal Relations: A Meta-Causal Perspectivecited by 2
  3. 2025Intervening to Learn and Compose Causally Disentangled Representationscited by 1
  4. 2024Disentangled Representations for Causal Cognitioncited by 0
  5. 2026Beyond identifiability: Learning causal representations with few environments and finite samplescited by 0