Research map / Optimization / Optimizers
Improve Single-Point Zeroth-Order Optimization Using High-Pass and Low-Pass Filters
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 a high-pass filter (residual feedback) to vanilla SZO (ablation) | SZO · Logistic regression, d=2 and N=200 | convergence speed | Helped |
| Add a high-pass filter (residual feedback) to vanilla SZO (ablation) | SZO · Logistic regression, d=2 and N=200 | estimation variance | Helped |
| Add a low-pass filter (momentum) to HF-SZO (ablation) | HF-SZO · Logistic regression, d=2 and N=200 | convergence speed | Helped |
| Set the high-pass parameter β to 1 (ablation) | HLF-SZO · Logistic regression, d=2 and N=200 | convergence speed | Helped |
| Set the high-pass parameter β to 1 (ablation) | HLF-SZO · Logistic regression, d=2 and N=200 | estimation variance | Helped |
| Add a low-pass filter (momentum) to residual-feedback SZO (ablation) | Residual-feedback SZO · Logistic regression, ridge regression, and Beale function | convergence speed | Helped |
| Add a low-pass filter (momentum) to residual-feedback SZO (ablation) | Residual-feedback SZO · LQR under different system noise levels | convergence speed | Helped |
| Replace HLF-SZO with vanilla SZO by removing both filters (ablation) | HLF-SZO · Logistic regression, d=2 and N=200 | convergence speed | Hurt |
| Replace HLF-SZO with two-point ZO | HLF-SZO · Logistic regression, ridge regression, Beale function, and LQR | convergence speed | No clear effect |
| Increase system noise | ZO methods · LQR | convergence speed | Hurt |
Closest papers by meaning
- Refining Adaptive Zeroth-Order Optimization at Ease (ICML 2025 · similarity 0.84)
- Double Variance Reduction: A Smoothing Trick for Composite Optimization Problems without First-Order Gradient (ICML 2024 · similarity 0.84)
- Adaptively Incorporating Directional Hints into Zeroth-Order Optimization (NeurIPS 2026 · similarity 0.83)
- Stochastic Zeroth-Order Optimization Under Heavy-Tailed Noise (NeurIPS 2026 · similarity 0.82)
- On the Optimal Construction of Unbiased Gradient Estimators for Zeroth-Order Optimization (NeurIPS 2025 · similarity 0.81)
- Trust the Direction, Search the Step: Zero-and-First-Order Methods for LLM Fine-Tuning (NeurIPS 2026 · similarity 0.81)
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.