mlsanityData to 7 Oct 2026

A Spectral Regularizer for Unsupervised Disentanglement

arXiv 1812.01161 ↗First posted 4 Dec 2018stat.ML cs.AI cs.LGcited by 30

Not on the map. Why?

Abstract (arXiv)

A generative model with a disentangled representation allows for independent control over different aspects of the output. Learning disentangled representations has been a recent topic of great interest, but it remains poorly understood. We show that even for GANs that do not possess disentangled representations, one can find curved trajectories in latent space over which local disentanglement occurs. These trajectories are found by iteratively following the leading right-singular vectors of the Jacobian of the generator with respect to its input. Based on this insight, we describe an efficient regularizer that aligns these vectors with the coordinate axes, and show that it can be used to induce disentangled representations in GANs, in a completely unsupervised manner.

Key facts. Cited by 30 papers in the map (data through 7 Oct 2026). First posted on arXiv 4 Dec 2018.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 7; 2022: 9; 2023: 5; 2024: 5; 2025: 3; 2026: 1.

30 papers.

Sort:

By area:

  1. 2021GAN Inversion: A Surveycited by 366
  2. 2023Understanding the Latent Space of Diffusion Models through the Lens of Riemannian Geometrycited by 121
  3. 2021Independent mechanism analysis, a new concept?cited by 96
  4. 2021Low-Rank Subspaces in GANscited by 52
  5. 2021WarpedGANSpace: Finding non-linear RBF paths in GAN latent spacecited by 48
  6. 2021Orthogonal Jacobian Regularization for Unsupervised Disentanglement in Image Generationcited by 35
  7. 2021EigenGAN: Layer-Wise Eigen-Learning for GANscited by 34
  8. 2023Unsupervised Discovery of Semantic Latent Directions in Diffusion Modelscited by 28
  9. 2021Do Not Escape From the Manifold: Discovering the Local Coordinates on the Latent Space of GANscited by 26
  10. 2022Symmetry-Based Representations for Artificial and Biological General Intelligencecited by 26
  11. 2022Region-Based Semantic Factorization in GANscited by 22
  12. 2025Personalized Image Generation with Deep Generative Models: A Decade Surveycited by 20
18 more
  1. 2022When are Post-hoc Conceptual Explanations Identifiable?cited by 14
  2. 2024Identifiable Causal Representation Learning: Unsupervised, Multi-View, and Multi-Environmentcited by 12
  3. 2023Generative Adversarial Networks Unlearningcited by 10
  4. 2023In-Domain GAN Inversion for Faithful Reconstruction and Editabilitycited by 5
  5. 2022Analyzing the Latent Space of GAN through Local Dimension Estimationcited by 3
  6. 2022Supervised Attribute Information Removal and Reconstruction for Image Manipulationcited by 3
  7. 2022Local Manifold Augmentation for Multiview Semantic Consistencycited by 3
  8. 2022Deep Curvilinear Editing: Commutative and Nonlinear Image Manipulation for Pretrained Deep Generative Modelcited by 3
  9. 2024Varying Manifolds in Diffusion: From Time-varying Geometries to Visual Saliencycited by 2
  10. 2026Mechanistically Eliciting Latent Behaviors in Language Modelscited by 2
  11. 2022A Survey on Leveraging Pre-trained Generative Adversarial Networks for Image Editing and Restorationcited by 1
  12. 2022Finding the global semantic representation in GAN through Frechet Meancited by 1
  13. 2024How Video Meetings Change Your Expressioncited by 1
  14. 2024(Deep) Generative Geodesicscited by 1
  15. 2025Be Tangential to Manifold: Discovering Riemannian Metric for Diffusion Modelscited by 1
  16. 2023Trained Latent Space Navigation to Prevent Lack of Photorealism in Generated Images on Style-based Modelscited by 0
  17. 2024Unsupervised Panoptic Interpretation of Latent Spaces in GANs Using Space-Filling Vector Quantizationcited by 0
  18. 2025Generative Fields: Uncovering Hierarchical Feature Control for StyleGAN via Inverted Receptive Fieldscited by 0