Research map / Optimization / Nonconvex and smooth optimization
A Unified Analysis of First-Order Methods for Smooth Games via Integral Quadratic Constraints
Guodong Zhang, Xuchan Bao, Laurent Lessard, Roger Grosse
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| add off-by-one pointwise IQCs to the sector IQCs when analyzing OG (ablation) | OG · smooth strongly-monotone games | certified convergence rate: rho >= 1 with sector IQCs only; improves bound by roughly a factor of 4 | Improved |
| tune OG step size much smaller than 1/4L under sector IQCs only | OG · time-varying vector field | iteration complexity | No clear effect |
| use proximal point analysis via IQC rather than prior bound | PPM · smooth strongly-monotone games | convergence rate rho^2: 1/(1 + eta m) -> 1/(1 + 2 eta m) | Improved |
| search over one-step-memory algorithm family parameters alpha, beta | one-step memory algorithms · kappa = 100 | convergence rate: optimal alpha = 1, beta = 0 | No clear effect |
| use negative momentum with tuned beta and step size | negative momentum · smooth strongly-monotone games, global | iteration complexity slope in kappa: slope of 1.5 | Improved |
Closest papers by meaning
- Classic but Everlasting: Traditional Gradient-Based Algorithms Converge Fast Even in Time-Varying Multi-Player Games (ICLR 2025 · similarity 0.89)
- Policy Gradient Methods Find the Nash Equilibrium in N-player General-sum Linear-quadratic Games (JMLR 2023 · similarity 0.88)
- Lyapunov Functions for First-Order Methods: Tight Automated Convergence Guarantees (ICML 2018 · similarity 0.88)
- Solving Zero-Sum Convex Markov Games (ICML 2025 · similarity 0.87)
- Optimistic Policy Gradient in Multi-Player Markov Games with a Single Controller: Convergence beyond the Minty Property (AAAI 2024 · similarity 0.86)
- Convergence of $\text{log}(1/\epsilon)$ for Gradient-Based Algorithms in Zero-Sum Games without the Condition Number: A Smoothed Analysis (NeurIPS 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.
