Research map / Optimization / Nonconvex and smooth optimization
Towards Better Understanding of Adaptive Gradient Algorithms in Generative Adversarial Nets
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use coordinate-wise adaptive steps based on cumulative gradient history instead of a fixed metric | OAdagrad versus OSG · Non-convex non-concave min-max optimization | iteration complexity: O(1/\epsilon^4) -> O(\epsilon^{-\frac{2}{1-\alpha}}) | Helped |
| use OAdagrad instead of OSG | WGAN-GP · CIFAR10 | Inception Score | Helped |
| use OAdagrad instead of Alternating Adam | WGAN-GP · CIFAR10 | Inception Score | Helped |
| use OAdagrad instead of Simultaneous Adam | Self-Attention GAN · ImageNet | Inception Score | Helped |
| use OAdagrad instead of Simultaneous Adam | Self-Attention GAN · ImageNet | Fréchet Inception Distance | Helped |
| use OAdagrad instead of OSG | GAN training · WGAN-GP on CIFAR10 | cumulative stochastic gradient growth degree: 0.2 | Mixed |
| use OAdagrad for GAN training where cumulative stochastic gradients grow slowly | OAdagrad · WGAN on LSUN Bedroom | cumulative stochastic gradient growth degree: 0.07 | Helped |
Closest papers by meaning
- Stability and Generalization of Stochastic Gradient Methods for Minimax Problems (ICML 2021 · similarity 0.89)
- A Convergent and Dimension-Independent Min-Max Optimization Algorithm (ICML 2022 · similarity 0.89)
- Tight Analysis of Extra-gradient and Optimistic Gradient Methods For Nonconvex Minimax Problems (NeurIPS 2022 · similarity 0.89)
- Non-asymptotic Analysis of Biased Adaptive Stochastic Approximation (NeurIPS 2024 · similarity 0.89)
- Complexity Lower Bounds of Adaptive Gradient Algorithms for Non-convex Stochastic Optimization under Relaxed Smoothness (ICLR 2025 · similarity 0.87)
- Adaptive Methods through the Lens of SDEs: Theoretical Insights on the Role of Noise (ICLR 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.