PAH learns per-task classifier heads from compact learned prototypes via a hypernetwork and uses dual knowledge distillation to keep accuracy high and forgetting near zero on Split-CIFAR100 and TinyImageNet.
Dark experience for general continual learning: a strong, simple baseline
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Prototype Augmented Hypernetworks for Continual Learning
PAH learns per-task classifier heads from compact learned prototypes via a hypernetwork and uses dual knowledge distillation to keep accuracy high and forgetting near zero on Split-CIFAR100 and TinyImageNet.