Research map / Continual learning / Online and replay-based
Learning Bayesian Sparse Networks With Full Experience Replay for Continual Learning
Check 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 SNCL instead of DER | continual learning · S-CIFAR-10 Class-IL, buffer 5120 | average accuracy: 83.81±0.33 -> 90.41±0.46 | Helped |
| add VBS to the DER baseline (ablation) | DER · S-CIFAR-10 Class-IL | accuracy: 61.93 -> 64.87 | Helped |
| add FER to the DER baseline (ablation) | DER · S-CIFAR-10 Class-IL | accuracy: 61.93 -> 65.85 | Helped |
| add LRS to the DER baseline (ablation) | DER · S-CIFAR-10 Class-IL | accuracy: 61.93 -> 63.07 | Helped |
| combine VBS and FER (ablation) | DER baseline · S-CIFAR-10 Class-IL | accuracy: 61.93 -> 66.02 | Helped |
| combine VBS and LRS (ablation) | DER baseline · S-CIFAR-10 Class-IL | accuracy: 61.93 -> 64.93 | Helped |
| combine FER and LRS (ablation) | DER baseline · S-CIFAR-10 Class-IL | accuracy: 61.93 -> 65.91 | Helped |
| combine VBS, FER, and LRS (ablation) | DER baseline · S-CIFAR-10 Class-IL | accuracy: 61.93 -> 66.16 | Helped |
| apply FER features at layer 1 rather than using the baseline (ablation) | DER baseline · P-MNIST Domain-IL | accuracy: 81.74 -> 82.98 | Helped |
| apply FER features at layer 2 rather than using the baseline (ablation) | DER baseline · P-MNIST Domain-IL | accuracy: 81.74 -> 85.27 | Helped |
| apply FER at both layers rather than only layer 1 (ablation) | DER baseline · P-MNIST Domain-IL | accuracy: 82.98 -> 85.53 | Helped |
| add VBS to FER applied only at layer 1 (ablation) | DER baseline · P-MNIST Domain-IL | accuracy: 82.98 -> 83.41 | Helped |
Closest papers by meaning
- Learning to Continually Learn with the Bayesian Principle (ICML 2024 · similarity 0.90)
- Layerwise Proximal Replay: A Proximal Point Method for Online Continual Learning (ICML 2024 · similarity 0.89)
- Task-aware Orthogonal Sparse Network for Exploring Shared Knowledge in Continual Learning (ICML 2024 · similarity 0.89)
- When to Learn and When to Stop: Quitting at the Optimal Time (Student Abstract) (AAAI 2025 · similarity 0.88)
- Mitigating Catastrophic Forgetting in Online Continual Learning by Modeling Previous Task Interrelations via Pareto Optimization (ICML 2024 · similarity 0.88)
- Do Your Best and Get Enough Rest for Continual Learning (CVPR 2025 · similarity 0.88)
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.