A new growth algorithm (DIRAD) and a prediction-validation framework (PREVAL) enable task-label-free continual learning on small MNIST tasks by escaping gradient conflicts and detecting new tasks.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Directed Structural Adaptation to Overcome Statistical Conflicts and Enable Continual Learning
A new growth algorithm (DIRAD) and a prediction-validation framework (PREVAL) enable task-label-free continual learning on small MNIST tasks by escaping gradient conflicts and detecting new tasks.