# Taming GANs with Lookahead-Minmax

arXiv: https://arxiv.org/abs/2006.14567 · HTML: https://mlsanity.com/paper/2006.14567/

## Key facts

- Cited by 16 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 25 Jun 2020.

## Built on by (most cited first, 16 of 16)

- [Near-optimal Local Convergence of Alternating Gradient Descent-Ascent for Minimax Optimization](https://mlsanity.com/paper/2102.09468/) (2021)
- [Decentralized Local Stochastic Extra-Gradient for Variational Inequalities](https://mlsanity.com/paper/2106.08315/) (2021)
- [Private GANs, Revisited](https://mlsanity.com/paper/2302.02936/) (2023)
- [SAN: Inducing Metrizability of GAN with Discriminative Normalized Linear Layer](https://mlsanity.com/paper/2301.12811/) (2023)
- [Last-Iterate Convergence of Saddle-Point Optimizers via High-Resolution Differential Equations](https://mlsanity.com/paper/2112.13826/) (2021)
- [Semantic Perturbations with Normalizing Flows for Improved Generalization](https://mlsanity.com/paper/2108.07958/) (2021)
- [Solving Constrained Variational Inequalities via a First-order Interior Point-based Method](https://mlsanity.com/paper/2206.10575/) (2022)
- [Stable Nonconvex-Nonconcave Training via Linear Interpolation](https://mlsanity.com/paper/2310.13459/) (2023)
- [Optimal Guarantees for Algorithmic Reproducibility and Gradient Complexity in Convex Optimization](https://mlsanity.com/paper/2310.17759/) (2023)
- [Multilayer Lookahead: a Nested Version of Lookahead](https://mlsanity.com/paper/2110.14254/) (2021)
- [Recursive Reasoning in Minimax Games: A Level k Gradient Play Method](https://mlsanity.com/paper/2210.16482/) (2022)
- [Stochastic Projective Splitting: Solving Saddle-Point Problems with Multiple Regularizers](https://mlsanity.com/paper/2106.13067/) (2021)
- [Training Generative Adversarial Networks with Adaptive Composite Gradient](https://mlsanity.com/paper/2111.05508/) (2021)
- [Decoupled SGDA for Games with Intermittent Strategy Communication](https://mlsanity.com/paper/2501.14652/) (2025)
- [Understanding Lookahead Dynamics Through Laplace Transform](https://mlsanity.com/paper/2506.13712/) (2025)
- [Frequency-Based Hyperparameter Selection in Games](https://mlsanity.com/paper/2601.18409/) (2026)

## Abstract (arXiv)

Generative Adversarial Networks are notoriously challenging to train. The underlying minmax optimization is highly susceptible to the variance of the stochastic gradient and the rotational component of the associated game vector field. To tackle these challenges, we propose the Lookahead algorithm for minmax optimization, originally developed for single objective minimization only. The backtracking step of our Lookahead-minmax naturally handles the rotational game dynamics, a property which was identified to be key for enabling gradient ascent descent methods to converge on challenging examples often analyzed in the literature. Moreover, it implicitly handles high variance without using large mini-batches, known to be essential for reaching state of the art performance. Experimental results on MNIST, SVHN, CIFAR-10, and ImageNet demonstrate a clear advantage of combining Lookahead-minmax with Adam or extragradient, in terms of performance and improved stability, for negligible memory and computational cost. Using 30-fold fewer parameters and 16-fold smaller minibatches we outperform the reported performance of the class-dependent BigGAN on CIFAR-10 by obtaining FID of 12.19 without using the class labels, bringing state-of-the-art GAN training within reach of common computational resources.