AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Designing to Forget: Deep Semi-parametric Models for Unlearning

Amber Yijia Zheng, Yu-Shan Tai, Raymond A. Yeh

CVPR 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Loss design and classifier theory · 0 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
add label-permutation augmentation (ablation)SPM generation · CIFAR-10 class-wise forgetting (Dog)Delta UA: 83.61 -> 0.04Improved
add label-permutation augmentation (ablation)SPM generation · CIFAR-10 class-wise forgetting (Horse)Delta UA: 60.89 -> 0.08Improved
add label-permutation augmentation (ablation)SPM generation · CIFAR-10 class-wise forgetting (Cat)Delta FID_R: 2.68 -> 0.50Improved
add label-permutation augmentation (ablation)SPM generation · CIFAR-10 class-wise forgetting (Automobile)FID_O: 2.88 -> 1.97Improved
use clustering (SPM-C) instead of retrieval (SPM-R) for the setSPM · ImageNet-1K classificationtop-1 accuracy: 59.9 -> 67.1Improved
use a larger inputted set at test time (10% -> 100%)SPM-C · ImageNet-1K classificationtop-1 accuracy: 66.6 -> 67.1Improved
use a larger inputted set |T| (10 -> 1024)SPM generation · CIFAR-10 class-conditioned generationFID: 7.71 -> 7.04Improved
use a larger inputted set |T| (10 -> 1024)SPM generation · CIFAR-10 class-conditioned generationruntime (s): 52 -> 1486Hurt
replace parametric ResNet18 with SPM-Cclassification · ImageNet-1K classificationtop-1 accuracy: 68.93 -> 67.1Hurt
use class-aware GNN as non-parametric moduleSPM · CIFAR-10 classificationaccuracy: 94.1 -> 94.4Improved
replace gradient-based unlearning (SalUn) with test-time deletion on SPMunlearning · ImageNet-1K 1-class unlearningPG_S: 0.42 -> 0.12Improved
replace gradient-based unlearning (SA) with test-time deletion on SPMunlearning · CIFAR-10 generation (Cat)FID_O: 32.54 -> 1.85Improved

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