Research map / Optimization / Implicit bias and generalization
A Sample Complexity Separation between Non-Convex and Convex Meta-Learning
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 a two-layer linear network with Reptile instead of convex linear regression | Initialization-based meta-learning · One-dimensional subspace meta-learning (new task) | sample complexity: \Omega\left(\frac{dr^2}{\epsilon}\right) -> \gO\left(\frac{r^2}{\epsilon}\right) | Helped |
| Use a two-layer linear network with RepLearn instead of convex linear regression | Initialization-based meta-learning · One-dimensional subspace meta-learning (new task) | sample complexity: \Omega\left(\frac{dr^2}{\epsilon}\right) -> \gO\left(\frac{r^2}{\epsilon}\right) | Helped |
| Analyze the representation-learning objective through its gradient-flow trajectory rather than arbitrary minimizers | RepLearn · One-dimensional subspace meta-learning (new task) | sample complexity: \Omega(d/\epsilon) -> \gO(1/\epsilon) | Helped |
Closest papers by meaning
- Provable Guarantees for Gradient-Based Meta-Learning (ICML 2019 · similarity 0.91)
- Adaptive Gradient-Based Meta-Learning Methods (NeurIPS 2019 · similarity 0.90)
- Provable Meta-Learning of Linear Representations (ICML 2021 · similarity 0.89)
- Learning To Learn Around A Common Mean (NeurIPS 2018 · similarity 0.89)
- On the Stability and Generalization of Meta-Learning (NeurIPS 2024 · similarity 0.87)
- Memory-Reduced Meta-Learning with Guaranteed Convergence (AAAI 2025 · similarity 0.87)
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.