Research map / Continual learning / Online and replay-based
Gradient Projection Memory for Continual Learning
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 GPM instead of OWM | Continual learning on PMNIST · PMNIST, 10 tasks | ACC (%): 90.71 -> 93.91 | Helped |
| Use GPM instead of HAT | Continual learning on CIFAR-100 · 10-Split CIFAR-100 | ACC (%): 72.06 -> 72.48 | Helped |
| Use GPM instead of HAT | Continual learning on miniImageNet · 20-Split miniImageNet | ACC (%): 59.78 -> 60.41 | Helped |
| Use GPM instead of HAT | Continual learning across diverse datasets · 5-Datasets | ACC (%): 91.32 -> 91.22 | Hurt |
| Use GPM instead of APD | Continual learning on related CIFAR-100 classes · 20-task CIFAR-100 Superclass | ACC (%): 56.81 -> 57.72 | Helped |
| Use GPM instead of DGDMN | Class-incremental digit learning · Digit dataset, class-incremental | ACC (%): 81.80 -> 70.67 | Hurt |
| Increase the representation threshold ε_th (ablation) | GPM · 10-Split CIFAR-100 | ACC (%) | Helped |
| Increase the representation threshold ε_th (ablation) | GPM · 10-Split CIFAR-100 | BWT | Helped |
| Use a subset of old-data replay with GPM | GPM in class-incremental learning · Digit dataset, class-incremental | ACC (%) | Helped |
Closest papers by meaning
- Continual Learning with Scaled Gradient Projection (AAAI 2023 · similarity 0.92)
- Convergence and Implicit Bias of Gradient Descent on Continual Linear Classification (ICLR 2025 · similarity 0.90)
- On the Theory of Continual Learning with Gradient Descent for Neural Networks (ICML 2026 · similarity 0.89)
- An Effective Dynamic Gradient Calibration Method for Continual Learning (ICML 2024 · similarity 0.88)
- Addressing Loss of Plasticity and Catastrophic Forgetting in Continual Learning (ICLR 2024 · similarity 0.87)
- Improving Continual Learning Performance and Efficiency with Auxiliary Classifiers (ICML 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.