Research map / Optimization / Optimizers
Efficient Riemannian Meta-Optimization by Implicit Differentiation
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 Riemannian implicit differentiation based on the final two inner-loop iterations instead of differentiating through the entire trajectory | Riemannian meta-optimization · PCA (MNIST), 250 inner-loop steps | training time (seconds): 4.88 × 102 -> 6.18 | Helped |
| Use Riemannian implicit differentiation based on the final two inner-loop iterations instead of differentiating through the entire trajectory | Riemannian meta-optimization · PCA (MNIST), 250 inner-loop steps | training memory (MB): 1.20 × 104 -> 4.55 × 103 | Helped |
| Use Riemannian implicit differentiation based on the final two inner-loop iterations instead of differentiating through the entire trajectory | Riemannian meta-optimization · Face Recognition (YaleB), 525 inner-loop steps | training time (seconds): 9.68 × 102 -> 2.72 | Helped |
| Use Riemannian implicit differentiation based on the final two inner-loop iterations instead of differentiating through the entire trajectory | Riemannian meta-optimization · Face Recognition (YaleB), 525 inner-loop steps | training memory (MB): 9.04 × 103 -> 2.60 × 103 | Helped |
| Use Riemannian implicit differentiation based on the final two inner-loop iterations instead of differentiating through the entire trajectory | Riemannian meta-optimization · Clustering (Kylberg texture) | training time: 3.53 × 102 -> 3.13 × 102 | Helped |
| Use Riemannian implicit differentiation based on the final two inner-loop iterations instead of differentiating through the entire trajectory | Riemannian meta-optimization · Clustering (Kylberg texture) | training memory: 8.73 × 102 -> 6.67 × 102 | Helped |
| Learn the optimizer with Riemannian implicit differentiation | Face Recognition (YaleB) · Face Recognition (YaleB) | accuracy (%): 90.2 -> 95.1 | Helped |
| Learn the optimizer with Riemannian implicit differentiation | Clustering (Kylberg texture) · Clustering (Kylberg texture) | accuracy (%): 85.2 -> 86.7 | Helped |
Closest papers by meaning
- Learning-Rate-Free Stochastic Optimization over Riemannian Manifolds (ICML 2024 · similarity 0.89)
- An Adaptive Algorithm for Bilevel Optimization on Riemannian Manifolds (NeurIPS 2025 · similarity 0.88)
- Riemannian Geometric-based Meta Learning (AAAI 2025 · similarity 0.88)
- Efficient Optimization with Orthogonality Constraint: a Randomized Riemannian Submanifold Method (ICML 2025 · similarity 0.88)
- Riemannian coordinate descent algorithms on matrix manifolds (ICML 2024 · similarity 0.87)
- Acceleration via silver step-size on Riemannian manifolds with applications to Wasserstein space (NeurIPS 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.