Towards Principled Methods for Training Generative Adversarial Networks
read the original abstract
The goal of this paper is not to introduce a single algorithm or method, but to make theoretical steps towards fully understanding the training dynamics of generative adversarial networks. In order to substantiate our theoretical analysis, we perform targeted experiments to verify our assumptions, illustrate our claims, and quantify the phenomena. This paper is divided into three sections. The first section introduces the problem at hand. The second section is dedicated to studying and proving rigorously the problems including instability and saturation that arize when training generative adversarial networks. The third section examines a practical and theoretically grounded direction towards solving these problems, while introducing new tools to study them.
This paper has not been read by Pith yet.
Forward citations
Cited by 12 Pith papers
-
VLTI/PIONIER imaging of post-AGB binaries. An INSPIRING hunt for inner rim substructures in circumbinary discs
High-resolution interferometric imaging of eight post-AGB circumbinary discs reveals diverse inner-rim substructures including azimuthal brightness enhancements and arc-like features not explained by inclination alone.
-
Causal Inference for Spatial Treatments
Develops design-based causal inference methods for spatial treatments using counterfactual candidate locations, extends double ML for spatial correlations, and applies to grocery store effects on foot traffic.
-
Causal Stability Selection
Causal stability selection identifies treatment effect modifiers with a non-asymptotic bound on expected false positives by integrating cross-fitted CATE estimation and stability selection.
-
A Dual Perspective on Synthetic Trajectory Generators: Utility Framework and Privacy Vulnerabilities
A new framework evaluates utility of synthetic mobility trajectories while a membership inference attack reveals privacy vulnerabilities in generative models thought to be safe.
-
Continuous Adversarial Flow Models
Continuous adversarial flow models replace MSE in flow matching with adversarial training via a discriminator, improving guidance-free FID on ImageNet from 8.26 to 3.63 for SiT and similar gains for JiT and text-to-im...
-
Physics-constrained generative machine learning-based high-resolution downscaling of Greenland's surface mass balance and surface temperature
A physics-constrained consistency model downscales Greenland SMB and surface temperature by a factor of 32 while preserving coarse-scale sums and outperforming interpolation on test metrics.
-
Copula & Marginal Flows: Disentangling the Marginal from its Joint
CM flows disentangle marginals from joints in normalizing flows to enable exact tail asymptotics and prior CDF assumptions via copula separation.
-
Demystifying MMD GANs
MMD GANs have unbiased critic gradients but biased generator gradients from sample-based learning, and the Kernel Inception Distance provides a practical new measure for GAN convergence and dynamic learning rate adaptation.
-
Finite-Time Analysis of MCTS in Continuous POMDP Planning
The paper proves finite-time probabilistic bounds on value estimates for MCTS in both discrete and continuous POMDPs and introduces Voro-POMCPOW with adaptive partitioning for guarantees.
-
Hard-Aware Fashion Attribute Classification
Presents HABP to emphasize hard samples during training and Deact to generate stable synthetic samples for rare attributes, outperforming prior methods on large-scale fashion datasets without extra supervision.
-
A Unified Measure-Theoretic View of Diffusion, Score-Based, and Flow Matching Generative Models
Diffusion, score-based, and flow matching models are unified as instances of learning time-dependent vector fields inducing marginal distributions governed by continuity and Fokker-Planck equations.
-
Cross-Machine Anomaly Detection Leveraging Pre-trained Time-series Model
A cross-machine anomaly detection framework disentangles MOMENT embeddings using random forests to create machine-invariant condition features that improve generalization to unseen machines on industrial data.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.