mlsanityData to 14 Oct 2026

On the Transfer of Disentangled Representations in Realistic Settings

arXiv 2010.14407 ↗First posted 27 Oct 2020cs.LG cs.AI cs.CVcited by 63

Not on the map. Why?

Abstract (arXiv)

Learning meaningful representations that disentangle the underlying structure of the data generating process is considered to be of key importance in machine learning. While disentangled representations were found to be useful for diverse tasks such as abstract reasoning and fair classification, their scalability and real-world impact remain questionable. We introduce a new high-resolution dataset with 1M simulated images and over 1,800 annotated real-world images of the same setup. In contrast to previous work, this new dataset exhibits correlations, a complex underlying structure, and allows to evaluate transfer to unseen simulated and real-world settings where the encoder i) remains in distribution or ii) is out of distribution. We propose new architectures in order to scale disentangled representation learning to realistic high-resolution settings and conduct a large-scale empirical study of disentangled representations on this dataset. We observe that disentanglement is a good predictor for out-of-distribution (OOD) task performance.

Key facts. Cited by 63 papers in the map (data through 14 Oct 2026). First posted on arXiv 27 Oct 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 17; 2022: 16; 2023: 14; 2024: 9; 2025: 4; 2026: 3.

63 papers.

Sort:

By area:

  1. 2021Towards Out-Of-Distribution Generalization: A Surveycited by 436
  2. 2021Contrastive Learning Inverts the Data Generating Processcited by 221
  3. 2021Towards Causal Representation Learningcited by 209
  4. 2021A Fine-Grained Analysis on Distribution Shiftcited by 171
  5. 2022Disentangled Representation Learningcited by 112
  6. 2024Entropy is not Enough for Test-Time Adaptation: From the Perspective of Disentangled Factorscited by 106
  7. 2022Out-Of-Distribution Generalization on Graphs: A Surveycited by 97
  8. 2021Towards Principled Disentanglement for Domain Generalizationcited by 91
  9. 2021Generalization and Robustness Implications in Object-Centric Learningcited by 81
  10. 2022Weakly Supervised Representation Learning with Sparse Perturbationscited by 65
  11. 2021Visual Representation Learning Does Not Generalize Strongly Within the Same Domaincited by 61
  12. 2021Identifiable Deep Generative Models via Sparse Decodingcited by 54
38 more
  1. 2022Mechanistic Mode Connectivitycited by 50
  2. 2021Learning Disentangled Representations in the Imaging Domaincited by 46
  3. 2022Synergies between Disentanglement and Sparsity: Generalization and Identifiability in Multi-Task Learningcited by 44
  4. 2022Compositional Generalization in Unsupervised Compositional Representation Learning: A Study on Disentanglement and Emergent Languagecited by 42
  5. 2022Robust and Efficient Medical Imaging with Self-Supervisioncited by 40
  6. 2022Disentanglement of Correlated Factors via Hausdorff Factorized Supportcited by 37
  7. 2021Sparse Interventions in Language Models with Differentiable Maskingcited by 34
  8. 2023Leveraging sparse and shared feature activations for disentangled representation learningcited by 32
  9. 2021On Causally Disentangled Representationscited by 22
  10. 2023Causal Triplet: An Open Challenge for Intervention-centric Causal Representation Learningcited by 22
  11. 2022Discrete Key-Value Bottleneckcited by 18
  12. 2022Pretraining in Deep Reinforcement Learning: A Surveycited by 18
  13. 2023Domain Generalisation via Risk Distribution Matchingcited by 14
  14. 2023Identifiable Contrastive Learning with Automatic Feature Importance Discoverycited by 14
  15. 2022Inductive Biases for Object-Centric Representations in the Presence of Complex Texturescited by 13
  16. 2025Does Data Scaling Lead to Visual Compositional Generalization?cited by 13
  17. 2021The Role of Pretrained Representations for the OOD Generalization of Reinforcement Learning Agentscited by 12
  18. 2021Exploring the Latent Space of Autoencoders with Interventional Assayscited by 10
  19. 2021Disentanglement and Generalization Under Correlation Shiftscited by 8
  20. 2022Overlooked Implications of the Reconstruction Loss for VAE Disentanglementcited by 7
  21. 2022Towards Unsupervised Content Disentanglement in Sentence Representations via Syntactic Rolescited by 5
  22. 2024Towards the Reusability and Compositionality of Causal Representationscited by 5
  23. 2022On Certifying and Improving Generalization to Unseen Domainscited by 4
  24. 2023Disentanglement Learning via Topologycited by 4
  25. 2023Sim-to-Real Causal Transfer: A Metric Learning Approach to Causally-Aware Interaction Representationscited by 4
  26. 2021Constellation: Learning relational abstractions over objects for compositional imaginationcited by 3
  27. 2023A Category-theoretical Meta-analysis of Definitions of Disentanglementcited by 3
  28. 2023On the Identifiability of Quantized Factorscited by 3
  29. 2023On the Fly Neural Style Smoothing for Risk-Averse Domain Generalizationcited by 3
  30. 2023The Acquisition of Physical Knowledge in Generative Neural Networkscited by 3
  31. 2024Graphs Generalization under Distribution Shiftscited by 3
  32. 2024Disentangled Representation Learning with the Gromov-Monge Gapcited by 3
  33. 2025Identifying Metric Structures of Deep Latent Variable Modelscited by 3
  34. 2023Enriching Disentanglement: From Logical Definitions to Quantitative Metricscited by 2
  35. 2024Toward A Causal Framework for Modeling Perceptioncited by 2
  36. 2024SGW-based Multi-Task Learning in Vision Taskscited by 2
  37. 2025Bridging Distribution Shift and AI Safety: Conceptual and Methodological Synergiescited by 2
  38. 2021Unsupervised Representation Disentanglement of Text: An Evaluation on Synthetic Datasetscited by 1

50 most cited of 63 All 63 (JSON)