Research map / Optimization / Optimizers
Learning to Learn by Zeroth-Order Oracle
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 |
|---|---|---|---|
| Use ZO-LSTM instead of hand-designed ZO optimizers | Black-box adversarial attack · MNIST | attack loss | Helped |
| Use ZO-LSTM instead of hand-designed ZO optimizers | Black-box adversarial attack · CIFAR-10 | attack loss | Helped |
| Use ZO-LSTM instead of hand-designed ZO optimizers | Stochastic binary classification · Synthetic binary classification | training loss | Helped |
| Add QueryRNN-guided Gaussian sampling (ablation) | ZO-LSTM · MNIST black-box attack | attack loss | Helped |
| Add QueryRNN-guided Gaussian sampling (ablation) | ZO-LSTM · CIFAR-10 black-box attack | attack loss | Helped |
| Add QueryRNN-guided Gaussian sampling (ablation) | ZO-LSTM · Synthetic binary classification | training loss | Helped |
| Add UpdateRNN-learned parameter updates (ablation) | ZO-LSTM · MNIST black-box attack | attack loss | Helped |
| Add UpdateRNN-learned parameter updates (ablation) | ZO-LSTM · CIFAR-10 black-box attack | attack loss | Helped |
| Add UpdateRNN-learned parameter updates (ablation) | ZO-LSTM · Synthetic binary classification | training loss | Helped |
| Replace QueryRNN sampling with GuidedES sampling (ablation) | ZO-LSTM · MNIST black-box attack | convergence speed | Helped |
| Replace QueryRNN sampling with GuidedES sampling (ablation) | ZO-LSTM · CIFAR-10 black-box attack | convergence speed | No clear effect |
| Replace QueryRNN sampling with GuidedES sampling (ablation) | ZO-LSTM · Synthetic binary classification | convergence speed | No clear effect |
Closest papers by meaning
- DeepZero: Scaling Up Zeroth-Order Optimization for Deep Model Training (ICLR 2024 · similarity 0.91)
- Refining Adaptive Zeroth-Order Optimization at Ease (ICML 2025 · similarity 0.88)
- Improving the Straight-Through Estimator with Zeroth-Order Information (NeurIPS 2025 · similarity 0.86)
- Zeroth-Order Optimization Meets Human Feedback: Provable Learning via Ranking Oracles (ICLR 2024 · similarity 0.86)
- Zeroth-Order Optimization Finds Flat Minima (NeurIPS 2025 · similarity 0.85)
- Natural Perturbations for Black-box Training of Neural Networks by Zeroth-Order Optimization (ICML 2025 · similarity 0.85)
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.