Research map / Continual learning / Online and replay-based
Coresets via Bilevel Optimization for Continual Learning and Streaming
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 |
|---|---|---|---|
| Replace iCaRL selection with bilevel coreset selection | Continual learning replay memory · PermMNIST | average test accuracy: 79.68 $\pm$ 0.41 -> 79.26 $\pm$ 0.43 | No clear effect |
| Replace iCaRL selection with bilevel coreset selection | Continual learning replay memory · SplitMNIST | average test accuracy: 93.99 $\pm$ 0.39 -> 95.87 $\pm$ 0.20 | Helped |
| Replace k-center embedding selection with bilevel coreset selection | Continual learning replay memory · SplitCIFAR-10 | average test accuracy: 36.91 $\pm$ 2.42 -> 37.60 $\pm$ 2.41 | No clear effect |
| Replace k-center summaries with bilevel coreset summaries in VCL | VCL · PermMNIST | average test accuracy: 85.33 $\pm$ 0.67 -> 86.18 $\pm$ 0.21 | Helped |
| Replace uniform summaries with bilevel coreset summaries in VCL | VCL · SplitMNIST | average test accuracy: 80.06 $\pm$ 2.19 -> 84.66 $\pm$ 0.66 | Helped |
| Replace reservoir sampling with merge-reduce bilevel coresets | Streaming replay memory · PermMNIST streaming | test accuracy: 73.21 $\pm$ 0.59 -> 74.44 $\pm$ 0.52 | Helped |
| Replace reservoir sampling with merge-reduce bilevel coresets | Streaming replay memory · SplitMNIST streaming | test accuracy: 90.72 $\pm$ 0.97 -> 92.59 $\pm$ 1.20 | Helped |
| Use bilevel coreset selection instead of CBRS | Imbalanced streaming replay memory · SplitMNIST | test accuracy: 89.71 $\pm$ 1.31 -> 92.30 $\pm$ 0.23 | Helped |
| Use bilevel coreset selection instead of CBRS | Imbalanced streaming replay memory · SplitCIFAR-10 | test accuracy: 32.25 $\pm$ 1.69 -> 33.98 $\pm$ 1.44 | Helped |
| Use bilevel coreset selection instead of reservoir sampling | Streaming replay memory · CIFAR-10 streaming | test accuracy | Hurt |
Closest papers by meaning
- Bilevel Coreset Selection in Continual Learning: A New Formulation and Algorithm (NeurIPS 2023 · similarity 0.91)
- Refined Coreset Selection: Towards Minimal Coreset Size under Model Performance Constraints (ICML 2024 · similarity 0.85)
- Bilevel Optimization under Unbounded Smoothness: A New Algorithm and Convergence Analysis (ICLR 2024 · similarity 0.83)
- Stable Coresets via Posterior Sampling: Aligning Induced and Full Loss Landscapes (NeurIPS 2025 · similarity 0.83)
- An Accelerated Algorithm for Stochastic Bilevel Optimization under Unbounded Smoothness (NeurIPS 2024 · similarity 0.83)
- A Nearly Optimal Single Loop Algorithm for Stochastic Bilevel Optimization under Unbounded Smoothness (ICML 2024 · similarity 0.83)
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.