Research map / Continual learning / Online and replay-based
Asymptotic analysis of shallow and deep forgetting in replay with neural collapse
Giulia Lanzillotta, Damiano Meier, Thomas Hofmann
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 |
|---|---|---|---|
| add a small replay buffer (about 5%) | Experience Replay · CIFAR100, ResNet | deep forgetting | Improved |
| increase replay buffer toward 100% | Experience Replay · CIFAR100, ResNet | shallow forgetting | Improved |
| initialize new heads with norm matched to previous heads (ablation) | Experience Replay (CIL/TIL) · CIFAR100 | forgetting and test accuracy | No clear effect |
| reduce feature dimension to 10 (bottleneck regime) (ablation) | Experience Replay · CIFAR100, ResNet, 5% replay | shallow vs deep forgetting gap: 52.27 -> 26.15 | No clear effect |
Closest papers by meaning
- Predicting the Susceptibility of Examples to Catastrophic Forgetting (ICML 2025 · similarity 0.91)
- Understanding the Forgetting of (Replay-based) Continual Learning via Feature Learning: Angle Matters (ICML 2025 · similarity 0.91)
- Measuring Representational Shifts in Continual Learning: A Linear Transformation Perspective (ICML 2025 · similarity 0.90)
- Layerwise Proximal Replay: A Proximal Point Method for Online Continual Learning (ICML 2024 · similarity 0.89)
- The Importance of Being Lazy: Scaling Limits of Continual Learning (ICML 2025 · similarity 0.89)
- Do Your Best and Get Enough Rest for Continual Learning (CVPR 2025 · similarity 0.89)
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.
