Research map / Optimization / Nonconvex and smooth optimization
Finding Differentially Private Second Order Stationary Points in Stochastic Minimax Optimization
Difei Xu, Youming Tao, Meng Ding, Chenglin Fan, Di Wang
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 |
|---|---|---|---|
| use DP-RGDA instead of Sto-SPIDER on explicit value function | DP-SOSP baselines · synthetic matrix sensing | Phi(x_399): 13.7046 -> 0.9119 | Improved |
| use DP-RGDA instead of Ada-DP-SPIDER | DP-SOSP baselines · synthetic matrix sensing | Phi(x_399): 8.0751 -> 0.9119 | Improved |
| use DP-RGDA instead of Sto-SPIDER | DP-SOSP baselines · synthetic matrix sensing | gradient norm: 3.4505 -> 0.4251 | Improved |
| use DP-RGDA instead of Ada-DP-SPIDER | DP-SOSP baselines · synthetic matrix sensing | gradient norm: 2.2753 -> 0.4251 | Improved |
| use DP-RGDA instead of Sto-SPIDER | DP-SOSP baselines · synthetic matrix sensing | lambda_min: -2.1862e-1 -> -4.8504e-2 | Improved |
| use PrivateDiff instead of DP-RGDA (first-order DP minimax baseline) | DP-RGDA · synthetic matrix sensing | Phi(x_399): 0.9119 -> 0.6546 | Hurt |
| use PrivateDiff instead of DP-RGDA | DP-RGDA · synthetic matrix sensing | lambda_min: -4.8504e-2 vs -4.3622e-2 | No clear effect |
Closest papers by meaning
- Second-Order Convergence in Private Stochastic Non-Convex Optimization (NeurIPS 2025 · similarity 0.93)
- Adaptive Batch Size for Privately Finding Second-Order Stationary Points (ICLR 2025 · similarity 0.91)
- Improved Sample Complexity for Private Nonsmooth Nonconvex Optimization (ICML 2025 · similarity 0.91)
- Private Stochastic Convex Optimization with Heavy Tails: Near-Optimality from Simple Reductions (NeurIPS 2024 · similarity 0.91)
- Differential Private Stochastic Optimization with Heavy-tailed Data: Towards Optimal Rates (AAAI 2025 · similarity 0.90)
- Faster Algorithms for User-Level Private Stochastic Convex Optimization (NeurIPS 2024 · similarity 0.89)
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.
