AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Model Sparsity Can Simplify Machine Unlearning

Jinghan Jia, Jiancheng Liu, Parikshit Ram, Yuguang Yao, Gaowen Liu, Yang Liu, Pranay Sharma, Sijia Liu

NeurIPS 2023 · Architectures and efficiency › Pruning and efficient tuning · approach: Layer-wise pruning of LLMs · 32 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
Prune to 95% sparsity with OMP before fine-tuningFT · CIFAR-10, ResNet-18, class-wise forgettingDisparity Ave. ↓: 25.78 -> 11.50Improved
Prune to 95% sparsity with OMP before influence unlearningIU · CIFAR-10, ResNet-18, class-wise forgettingDisparity Ave. ↓: 5.41 -> 1.10Improved
Prune to 95% sparsity with OMP before fine-tuningFT · CIFAR-10, ResNet-18, random data forgettingDisparity Ave. ↓: 2.74 -> 1.92Improved
Prune to 95% sparsity with OMP before influence unlearningIU · CIFAR-10, ResNet-18, random data forgettingDisparity Ave. ↓: 4.07 -> 3.08Improved
Add constant ℓ1 sparsity regularization during fine-tuning (ablation)MUSparse · CIFAR-10, ResNet-18, random data forgetting (10%)UA: 6.60 (1.19)Mixed
Increase ℓ1 regularization linearly during fine-tuning (ablation)MUSparse · CIFAR-10, ResNet-18, random data forgetting (10%)UA: 3.80 (1.61)Mixed
Decrease ℓ1 regularization linearly during fine-tuning (ablation)MUSparse · CIFAR-10, ResNet-18, random data forgetting (10%)UA: 5.35 (0.06)Improved
Use SynFlow pruning instead of OMP before fine-tuning (ablation)FT · CIFAR-10, ResNet-18, class-wise forgettingUA and MIAFMixed
Use IMP pruning instead of OMP before fine-tuning (ablation)FT · CIFAR-10, ResNet-18, class-wise forgettingUA and MIAFHurt
Prune to 95% sparsity before approximate unlearningGA · CIFAR-10, ResNet-18, class-wise forgettingDisparity Ave. ↓: 7.12 -> 6.41Improved
Prune to 95% sparsity before approximate unlearningFF · CIFAR-10, ResNet-18, class-wise forgettingDisparity Ave. ↓: 5.32 -> 1.98Improved
Prune to 95% sparsity before approximate unlearningGA · CIFAR-10, ResNet-18, random data forgettingDisparity Ave. ↓: 4.26 -> 3.05Improved

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