# Is Generator Conditioning Causally Related to GAN Performance?

arXiv: https://arxiv.org/abs/1802.08768 · HTML: https://mlsanity.com/paper/1802.08768/

## Key facts

- Cited by 18 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 23 Feb 2018.

## Built on by (most cited first, 18 of 18)

- [Large Scale Image Completion via Co-Modulated Generative Adversarial Networks](https://mlsanity.com/paper/2103.10428/) (2021)
- [Smooth Diffusion: Crafting Smooth Latent Spaces in Diffusion Models](https://mlsanity.com/paper/2312.04410/) (2023)
- [3D generation on ImageNet](https://mlsanity.com/paper/2303.01416/) (2023)
- [Can Push-forward Generative Models Fit Multimodal Distributions?](https://mlsanity.com/paper/2206.14476/) (2022)
- [Orthogonal Jacobian Regularization for Unsupervised Disentanglement in Image Generation](https://mlsanity.com/paper/2108.07668/) (2021)
- [Convolutional Normalization: Improving Deep Convolutional Network Robustness and Training](https://mlsanity.com/paper/2103.00673/) (2021)
- [LinkGAN: Linking GAN Latents to Pixels for Controllable Image Synthesis](https://mlsanity.com/paper/2301.04604/) (2023)
- [Adversarial sampling of unknown and high-dimensional conditional distributions](https://mlsanity.com/paper/2111.05962/) (2021)
- [Analyzing and Improving Optimal-Transport-based Adversarial Networks](https://mlsanity.com/paper/2310.02611/) (2023)
- [FedSyn: Synthetic Data Generation using Federated Learning](https://mlsanity.com/paper/2203.05931/) (2022)
- [A Brief Introduction to Generative Models](https://mlsanity.com/paper/2103.00265/) (2021)
- [A Survey on Bridging EEG Signals and Generative AI: From Image and Text to Beyond](https://mlsanity.com/paper/2502.12048/) (2025)
- [Diverse Shape Completion via Style Modulated Generative Adversarial Networks](https://mlsanity.com/paper/2311.11184/) (2023)
- [Detecting Anomalies using Generative Adversarial Networks on Images](https://mlsanity.com/paper/2211.13808/) (2022)
- [Scaling Quantum Machine Learning without Tricks: Full-Resolution and Diverse Image Generation](https://mlsanity.com/paper/2603.00233/) (2026)
- [RCC-GAN: Regularized Compound Conditional GAN for Large-Scale Tabular Data Synthesis](https://mlsanity.com/paper/2205.11693/) (2022)
- [Wavelet Prior Attention Learning in Axial Inpainting Network](https://mlsanity.com/paper/2206.03113/) (2022)
- [MediQ-GAN: Quantum-Inspired GAN for High Resolution Medical Image Generation](https://mlsanity.com/paper/2506.21015/) (2025)

## Abstract (arXiv)

Recent work (Pennington et al, 2017) suggests that controlling the entire distribution of Jacobian singular values is an important design consideration in deep learning. Motivated by this, we study the distribution of singular values of the Jacobian of the generator in Generative Adversarial Networks (GANs). We find that this Jacobian generally becomes ill-conditioned at the beginning of training. Moreover, we find that the average (with z from p(z)) conditioning of the generator is highly predictive of two other ad-hoc metrics for measuring the 'quality' of trained GANs: the Inception Score and the Frechet Inception Distance (FID). We test the hypothesis that this relationship is causal by proposing a 'regularization' technique (called Jacobian Clamping) that softly penalizes the condition number of the generator Jacobian. Jacobian Clamping improves the mean Inception Score and the mean FID for GANs trained on several datasets. It also greatly reduces inter-run variance of the aforementioned scores, addressing (at least partially) one of the main criticisms of GANs.