Gradient Concentration, Not Weight Saliency, Explains Representation-Level Class Unlearning
Machine unlearning aims to remove the influence of specific training data while preserving model utility. Many state-of-the-art approaches pursue this goal by restricting the forgetting update to a subset of parameters selected through gradient-based saliency. Although such methods are widely adopted, the actual contribution of saliency-based weight selection to representation-level forgetting remains unclear. In this work, we perform the first controlled ablation of the saliency masking mechanism used by SalUn. Using a matched-compute experimental design on CIFAR-10 and CIFAR-100 with ResNet-
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- PossiblePossibly related (embedding) · 50%Breakthrough in long-context efficiency announced →
- LinkedLinked via arxiv author · 85%Billel Habbati →
“Gradient Concentration, Not Weight Saliency, Explains Representation-Level Class Unlearning”
- LinkedLinked via arxiv author · 85%Alessio Merlo →
“Gradient Concentration, Not Weight Saliency, Explains Representation-Level Class Unlearning”
- LinkedLinked via arxiv author · 85%Luca Verderame →
“Gradient Concentration, Not Weight Saliency, Explains Representation-Level Class Unlearning”
- LinkedLinked via arxiv author · 85%Meriem Guerar →
“Gradient Concentration, Not Weight Saliency, Explains Representation-Level Class Unlearning”