Research map / Optimization / Nonconvex and smooth optimization
Weaker MVI Condition: Extragradient Methods with Multi-Step Exploration
Yifeng Fan, Yongqiang Li, Bo Chen
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes a multi-step extragradient framework for nonconvex-nonconcave min-max problems that works under a weaker MVI (minty variational inequality) condition.
Problem
Solving nonconvex-nonconcave min-max optimization problems (e.g., GAN training), where standard convex-concave methods and existing extragradient analyses face convergence difficulties under restrictive assumptions.
Approach
Extends the extragradient algorithm to a multi-step exploration scheme and proves convergence under a weaker Minty variational inequality condition than previously assumed.
Experimental setup
Builds on: extragradient
Closest papers by meaning
- Solving stochastic weak Minty variational inequalities without increasing batch size (ICLR 2023 · similarity 0.90)
- First-order Convergence Theory for Weakly-Convex-Weakly-Concave Min-max Problems (JMLR 2021 · similarity 0.90)
- Escaping limit cycles: Global convergence for constrained nonconvex-nonconcave minimax problems (ICLR 2022 · similarity 0.89)
- Revisiting Inexact Fixed-Point Iterations for Min-Max Problems: Stochasticity and Structured Nonconvexity (ICML 2024 · similarity 0.87)
- Two-timescale Extragradient for Finding Local Minimax Points (ICLR 2024 · similarity 0.86)
- Double-Step Alternating Extragradient with Increasing Timescale Separation for Finding Local Minimax Points: Provable Improvements (ICML 2024 · similarity 0.86)
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.
