English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

With evolving data regulations, machine unlearning (MU) has become animportant tool for fostering trust and safety in today's AI models. However,existing MU methods focusing on data and/or weight perspectives often grapplewith limitations in unlearning accuracy, stability, and cross-domainapplicability. To address these challenges, we introduce the concept of 'weightsaliency' in MU, drawing parallels with input saliency in model explanation.This innovation directs MU's attention toward specific model weights ratherthan the entire model, improving effectiveness and efficiency. The resultantmethod that we call saliency unlearning (SalUn) narrows the performance gapwith 'exact' unlearning (model retraining from scratch after removing theforgetting dataset). To the best of our knowledge, SalUn is the firstprincipled MU approach adaptable enough to effectively erase the influence offorgetting data, classes, or concepts in both image classification andgeneration. For example, SalUn yields a stability advantage in high-variancerandom data forgetting, e.g., with a 0.2% gap compared to exact unlearning onthe CIFAR-10 dataset. Moreover, in preventing conditional diffusion models fromgenerating harmful images, SalUn achieves nearly 100% unlearning accuracy,outperforming current state-of-the-art baselines like Erased Stable Diffusionand Forget-Me-Not.

SalUn: Empowering Machine Unlearning via Gradient-based Weight Saliency in Both Image Classification and Generation