AI Paper Atlas

Research map / Federated learning and privacy / Privacy attacks and secure learning

Mitigating Privacy Risk via Forget Set-Free Unlearning

Aviraj Newatia, Michael Cooper, Viet Nguyen, Rahul G Krishnan

ICLR 2026 · Federated learning and privacy › Privacy attacks and secure learning · approach: Encryption-based private learning · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use Reload instead of GAunlearning methods · CIFAR-100 forget 10% randomdelta forget accuracy: 18.77 -> 0.30Improved
Reload (without forget set) vs forget-set methodsReload · CIFAR-100 forget 10% randomdelta forget MIA: 0.21 -> 0.01Improved
add selective re-initialisation after one ascent step (ablation)Reload · SVHN class 8 forgettingpredicted classImproved
Reload on TOFU 1% forgettingLlama-2-7B / Phi-1.5 · TOFU 1%model utility: +0.0748Improved
Reload on TOFU 5% and 10% forgettingReload · TOFU 5%/10%model utilityHurt
finetune on subsets of retain set (ablation)Reload · unlearningperformanceHurt
quantize cached gradients (ablation)Reload · unlearningperformanceNo clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy