All papers
Investigating the Role of Samples in Catastrophic Forgetting
Simon Bührer, Deep Learning course project, ETH Zürich, March 2025
Abstract
A study of how the choice of replayed examples affects catastrophic forgetting in continual learning. It compares three ways of deciding which samples to keep in the replay buffer: confidence-based learning-speed estimation, weighted prioritisation, and sensitivity-aware sampling based on the Memory-Perturbation Equation. On CIFAR-10 with a ResNet-18, all three end up close to the Goldilocks baseline without clearly beating it. The confidence-based scores track Goldilocks well (Pearson r up to 0.83) but use more memory.
Tags
- Continual Learning
- Replay Memory
- Catastrophic Forgetting