Stochastic Gradient Descent Ascent is Suboptimal for Nonconvex-P Min-Max Games
stat.ML, cs.GT, cs.LG
Submitted: 2026-10-06
Updated: 2026-10-09
Sources
- The Complexity of Min-Max Optimization for Quadratic Polynomials
- Accelerated Stochastic Min-Max Optimization Based on Bias-corrected Momentum
- Faster Stochastic Algorithms for Minimax Optimization under Polyak--{\L}ojasiewicz Conditions
- Generative Adversarial Reasoner: Enhancing LLM Reasoning with Adversarial Reinforcement Learning
- PL conditions do not guarantee convergence of gradient descent-ascent dynamics
- Negative Stepsizes Make Gradient-Descent-Ascent Converge
- Certifying Some Distributional Robustness with Principled Adversarial Training
- Efficient Algorithms for Smooth Minimax Optimization
- Robust LLM Alignment via Distributionally Robust Direct Preference Optimization
Related papers
- Behavior of prediction performance metrics with rare events
- Optimal Estimation of Generic Dynamics by Path-Dependent Neural Jump ODEs
- A Posterior-Dynamics Framework for Imaging Inverse Problems with Pretrained Diffusion Priors
- Efficient quadratic entropy with distance sketches
- Embedding-Bias in Conditional Independence Testing
- From Log-Odds to Shapley Values: An Explanatory Geometry for the Weighted Naive Bayes Classifier