mlsanityData to 7 Oct 2026

Single-level Adversarial Data Synthesis based on Neural Tangent Kernels

arXiv 2204.04090 ↗First posted 8 Apr 2022cs.LGcited by 2

Not on the map. Why?

Abstract (arXiv)

Abstract Generative adversarial networks (GANs) have achieved impressive performance in data synthesis and have driven the development of many applications. However, GANs are known to be hard to train due to their bilevel objective, which leads to the problems of convergence, mode collapse, and gradient vanishing. In this paper, we propose a new generative model called the generative adversarial NTK (GA-NTK) that has a single-level objective. The GA-NTK keeps the spirit of adversarial learning (which helps generate plausible data) while avoiding the training difficulties of GANs. This is done by modeling the discriminator as a Gaussian process with a neural tangent kernel (NTK-GP) whose training dynamics can be completely described by a closed-form formula. We analyze the convergence behavior of GA-NTK trained by gradient descent and give some sufficient conditions for convergence. We also conduct extensive experiments to study the advantages and limitations of GA-NTK and propose some techniques that make GA-NTK more practical.

Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 60 of its 80 references that are in the map. First posted on arXiv 8 Apr 2022.

Builds on

60 of 80 references.

References · 60

  1. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  2. 2022Generative Adversarial Networkscited by 5,603
  3. 2014Deep Learning Face Attributes in the Wildcited by 3,679
  4. 2019Analyzing and Improving the Image Quality of StyleGANcited by 3,580
  5. 2016Improved Techniques for Training GANscited by 2,938
  6. 2018Large Scale GAN Training for High Fidelity Natural Image Synthesiscited by 2,609
  7. 2018Neural Tangent Kernel: Convergence and Generalization in Neural Networkscited by 2,597
  8. 2017Improved Training of Wasserstein GANscited by 2,223
  9. 2015Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networkscited by 2,048
  10. 2016Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Networkcited by 1,898
  11. 2018Spectral Normalization for Generative Adversarial Networkscited by 1,618
  12. 2017High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANscited by 1,176
38 more
  1. 2016Least Squares Generative Adversarial Networkscited by 1,000
  2. 2019Adversarial Examples Are Not Bugs, They Are Featurescited by 950
  3. 2016InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Netscited by 858
  4. 2019Wide Neural Networks of Any Depth Evolve as Linear Models Under Gradient Descentcited by 732
  5. 2018Which Training Methods for GANs do actually Converge?cited by 672
  6. 2018On Lazy Training in Differentiable Programmingcited by 671
  7. 2017Deep Neural Networks as Gaussian Processescited by 617
  8. 2019On Exact Computation with an Infinitely Wide Neural Netcited by 572
  9. 2017Towards Principled Methods for Training Generative Adversarial Networkscited by 460
  10. 2016Generating Videos with Scene Dynamicscited by 376
  11. 2018Gaussian Process Behaviour in Wide Deep Neural Networkscited by 314
  12. 2016On the Expressive Power of Deep Neural Networkscited by 302
  13. 2016NIPS 2016 Tutorial: Generative Adversarial Networkscited by 284
  14. 2017Are GANs Created Equal? A Large-Scale Studycited by 267
  15. 2018Improving Unsupervised Defect Segmentation by Applying Structural Similarity to Autoencoderscited by 263
  16. 2016Exponential expressivity in deep neural networks through transient chaoscited by 260
  17. 2015Generative Moment Matching Networkscited by 252
  18. 2017MMD GAN: Towards Deeper Understanding of Moment Matching Networkcited by 251
  19. 2017Training GANs with Optimismcited by 217
  20. 2018Sorting out Lipschitz function approximationcited by 211
  21. 2016Unrolled Generative Adversarial Networkscited by 208
  22. 2017VEEGAN: Reducing Mode Collapse in GANs using Implicit Variational Learningcited by 199
  23. 2016Deep Information Propagationcited by 194
  24. 2019On the Inductive Bias of Neural Tangent Kernelscited by 177
  25. 2018Bayesian Deep Convolutional Networks with Many Channels are Gaussian Processescited by 164
  26. 2020Finite Versus Infinite Neural Networks: an Empirical Studycited by 156
  27. 2019Neural Tangents: Fast and Easy Infinite Neural Networks in Pythoncited by 149
  28. 2019Tensor Programs I: Wide Feedforward or Recurrent Neural Networks of Any Architecture are Gaussian Processescited by 148
  29. 2019Scaling Limits of Wide Neural Networks with Weight Sharing: Gaussian Process Behavior, Gradient Independence, and Neural Tangent Kernel Derivationcited by 146
  30. 2019A Unified Analysis of Extra-gradient and Optimistic Gradient Methods for Saddle Point Problems: Proximal Point Approachcited by 142
  31. 2019Mode Seeking Generative Adversarial Networks for Diverse Image Synthesiscited by 138
  32. 2018Deep Convolutional Networks as shallow Gaussian Processescited by 137
  33. 2016Mode Regularized Generative Adversarial Networkscited by 123
  34. 2020Infinite attention: NNGP and NTK for deep attention networkscited by 117
  35. 2019Harnessing the Power of Infinitely Wide Deep Nets on Small-data Taskscited by 87
  36. 2020On the Similarity between the Laplace and Neural Tangent Kernelscited by 76
  37. 2019Efficient Algorithms for Smooth Minimax Optimizationcited by 75
  38. 2020The Recurrent Neural Tangent Kernelcited by 67

50 most cited of 60 All 60 (JSON)

Built on by

Cited by year

Citing papers per year: 2023: 2.

2 papers.

  1. 2023Differentially Private Neural Tangent Kernels for Privacy-Preserving Data Generationcited by 15
  2. 2023Data Interpolants -- That's What Discriminators in Higher-order Gradient-regularized GANs Arecited by 2