Research map / Learning theory and algorithms / Sample complexity
Bilevel Optimization with a Lower-level Contraction: Optimal Sample Complexity without Warm-Start
Riccardo Grazzi, Massimiliano Pontil, Saverio Salzo
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 warm-start on the lower-level problem (ablation) | BSGM · Equilibrium models on MNIST | proximal gradient map norm | Improved |
| Use warm-start on the lower-level problem (ablation) | BSGM · Equilibrium models on MNIST | train and test accuracy | No clear effect |
| Reduce the upper-level mini-batch size (ablation) | BSGM · Equilibrium models on MNIST | performance by number of epochs | Improved |
| Use warm-start by storing previously sampled task solutions (ablation) | BSGM · 5-way 5-shot Mini-Imagenet | reported metrics | No clear effect |
| Use warm-start by storing previously sampled task solutions (ablation) | BSGM · 5-way 5-shot Mini-Imagenet | memory cost: 0.8 GB | Hurt |
| Use warm-start (ablation) | Bilevel optimization methods · Data poisoning on MNIST | performance | Mixed |
Closest papers by meaning
- Provably Faster Algorithms for Bilevel Optimization via Without-Replacement Sampling (NeurIPS 2024 · similarity 0.89)
- Penalty-Based First-Order Methods for Bilevel Optimization with Minimax and Constrained Lower-Level Problems (NeurIPS 2026 · similarity 0.88)
- SPABA: A Single-Loop and Probabilistic Stochastic Bilevel Algorithm Achieving Optimal Sample Complexity (ICML 2024 · similarity 0.87)
- TSP: A Two-Sided Smoothed Primal-Dual Method for Nonconvex Bilevel Optimization (ICML 2025 · similarity 0.87)
- On the Sample Complexity Bounds of Bilevel Reinforcement Learning (NeurIPS 2025 · similarity 0.87)
- An Accelerated Gradient Method for Convex Smooth Simple Bilevel Optimization (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.
