Provides the first theoretical formulation and bounds for certified unlearning in continual learning, decomposing excess risk and adapting gradient and Hessian methods with a low-storage hybrid.
A unified gradient-based framework for task-agnostic contin- ual learning-unlearning.arXiv preprint arXiv:2505.15178
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
fields
cs.LG 3years
2026 3verdicts
UNVERDICTED 3representative citing papers
SAFER is a continual unlearning method that prevents progressive accuracy loss on retain data and reversal of forgetting by enforcing representation stability and negative logit margins.
BID-LoRA uses bi-directional low-rank adapters with retain/new/unlearn pathways and escape unlearning to enable continual learning and unlearning while minimizing knowledge leakage and parameter updates.
citing papers explorer
-
The Forgetting-Retention Dilemma: Certified Unlearning Theory in Continual Learning
Provides the first theoretical formulation and bounds for certified unlearning in continual learning, decomposing excess risk and adapting gradient and Hessian methods with a low-storage hybrid.
-
Robust Continual Unlearning against Knowledge Erosion and Forgetting Reversal
SAFER is a continual unlearning method that prevents progressive accuracy loss on retain data and reversal of forgetting by enforcing representation stability and negative logit margins.
-
BID-LoRA: A Parameter-Efficient Framework for Continual Learning and Unlearning
BID-LoRA uses bi-directional low-rank adapters with retain/new/unlearn pathways and escape unlearning to enable continual learning and unlearning while minimizing knowledge leakage and parameter updates.