CODE-CL combines conceptor matrices with a learnable shared-direction projection to reduce catastrophic forgetting and improve forward transfer in task-incremental learning.
Adversarial Continual Learn- ing
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
-
CODE-CL: Conceptor-Based Gradient Projection for Deep Continual Learning
CODE-CL combines conceptor matrices with a learnable shared-direction projection to reduce catastrophic forgetting and improve forward transfer in task-incremental learning.