Research map / Optimization / Optimizers
On the Convergence Theory for Hessian-Free Bilevel Algorithms
Daouda Sow, Kaiyi Ji, Yingbin Liang
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 |
|---|---|---|---|
| PZOBO instead of ProtoNet | PZOBO · miniImageNet ResNet-12 | hours to 67% accuracy: 3.5 -> 2.0 | Improved |
| PZOBO instead of ProtoNet | PZOBO · miniImageNet ResNet-12 | hours to 69% accuracy: 4.4 -> 2.8 | Improved |
| PZOBO instead of MAML | PZOBO · miniImageNet ResNet-12 | hours to 67% accuracy: 9.7 -> 2.0 | Improved |
| PZOBO instead of ANIL | PZOBO · miniImageNet ResNet-12 | hours to 67% accuracy: 6.3 -> 2.0 | Improved |
| PZOBO instead of MetaOptNet | PZOBO · miniImageNet ResNet-12 | hours to 69% accuracy: 20+ -> 2.8 | Improved |
| PZOBO-S instead of other bilevel methods | PZOBO-S · MNIST deep HR | accuracy: 0.93 -> 0.98 | Improved |
| estimate only the response Jacobian (PZOBO) instead of full hypergradient (HOZOG) | PZOBO · shallow HR two-layer net | convergence | Improved |
| vary number N of inner GD steps (ablation) | PZOBO · shallow HR | convergence | No clear effect |
| vary number N of inner SGD steps (ablation) | PZOBO-S · MNIST deep HR | convergence | No clear effect |
| PZOBO instead of HOZOG | PZOBO · 20 Newsgroup hyperparameter optimization | convergence | Improved |
| increase network size from CNN4 to ResNet-12 | PZOBO · miniImageNet | advantage over MAML and ANIL | Improved |
Closest papers by meaning
- Tuning-Free Bilevel Optimization: New Algorithms and Convergence Analysis (ICLR 2025 · similarity 0.92)
- Efficient Curvature-Aware Hypergradient Approximation for Bilevel Optimization (ICML 2025 · similarity 0.92)
- qNBO: quasi-Newton Meets Bilevel Optimization (ICLR 2025 · similarity 0.92)
- Optimal Hessian/Jacobian-Free Nonconvex-PL Bilevel Optimization (ICML 2024 · similarity 0.91)
- Bilevel Optimization under Unbounded Smoothness: A New Algorithm and Convergence Analysis (ICLR 2024 · similarity 0.91)
- A Nearly Optimal Single Loop Algorithm for Stochastic Bilevel Optimization under Unbounded Smoothness (ICML 2024 · similarity 0.90)
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.
