# Tempered Adversarial Networks

arXiv: https://arxiv.org/abs/1802.04374 · HTML: https://mlsanity.com/paper/1802.04374/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 12 Feb 2018.

## Built on by (most cited first, 6 of 6)

- [Diffusion-Based Representation Learning](https://mlsanity.com/paper/2105.14257/) (2021)
- [From Points to Functions: Infinite-dimensional Representations in Diffusion Models](https://mlsanity.com/paper/2210.13774/) (2022)
- [Single-level Adversarial Data Synthesis based on Neural Tangent Kernels](https://mlsanity.com/paper/2204.04090/) (2022)
- [Generative Adversarial Networks via a Composite Annealing of Noise and Diffusion](https://mlsanity.com/paper/2105.00220/) (2021)
- [Boltzmann Tuning of Generative Models](https://mlsanity.com/paper/2104.05252/) (2021)
- [Diffusion Based Causal Representation Learning](https://mlsanity.com/paper/2311.05421/) (2023)

## Abstract (arXiv)

Generative adversarial networks (GANs) have been shown to produce realistic samples from high-dimensional distributions, but training them is considered hard. A possible explanation for training instabilities is the inherent imbalance between the networks: While the discriminator is trained directly on both real and fake samples, the generator only has control over the fake samples it produces since the real data distribution is fixed by the choice of a given dataset. We propose a simple modification that gives the generator control over the real samples which leads to a tempered learning process for both generator and discriminator. The real data distribution passes through a lens before being revealed to the discriminator, balancing the generator and discriminator by gradually revealing more detailed features necessary to produce high-quality results. The proposed module automatically adjusts the learning process to the current strength of the networks, yet is generic and easy to add to any GAN variant. In a number of experiments, we show that this can improve quality, stability and/or convergence speed across a range of different GAN architectures (DCGAN, LSGAN, WGAN-GP).