AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Dynamic Sparsity Is Channel-Level Sparsity Learner

Lu Yin, Gen Li, Meng Fang, Li Shen, Tianjin Huang, Zhangyang Wang, Vlado Menkovski, Xiaolong Ma, Mykola Pechenizkiy, Shiwei Liu

NeurIPS 2023 · Architectures and efficiency › Pruning and efficient tuning · approach: Sparse training from scratch · 3 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 global parameter exploration to RigL (Chase Sc=0) (ablation)RigL · ImageNet, ResNet-50, 80% sparsitytop-1 accuracy: 75.1 -> 75.87Improved
add soft memory bound (ablation)Chase/RigL · CIFAR-100, ResNet-50, 98% sparsityaccuracy: +1.26%Improved
add global parameter exploration on top of soft memory bound (ablation)Chase · CIFAR-10/100, ResNet-50 and VGG-19accuracyImproved
add gradual amenable channel pruning (50% channels) to RigL (ablation)RigL · CIFAR-100, ResNet-50, 98% sparsityaccuracy: 74.84 -> 74.92Improved
add gradual amenable channel pruning (50% channels) to SET (ablation)SET · CIFAR-100, ResNet-50, 98% sparsityaccuracy: 73.21 -> 73.86Improved
add gradual amenable channel pruning to RigL (ablation)RigL · CIFAR-100, ResNet-50, 90% sparsityaccuracy: 77.18 -> 76.86Hurt
use UMM instead of random pruning for one-shot channel pruningRigL · CIFAR-100, ResNet-50, 30% channel sparsityaccuracy: 3.79 -> 71.77Improved
prune with reversed UMM metric instead of UMMRigL · CIFAR-100, ResNet-50, 10% channel sparsityaccuracy: 76.88 -> 3.18Hurt
add Chase to SNIP (static sparse training)SNIP · CIFAR-100, ResNet-50, 90% sparsityaccuracy: 73.14 -> 73.44Improved
add Chase to GraSP (static sparse training)GraSP · CIFAR-100, ResNet-50, 98% sparsityaccuracy: 62.12 -> 75.74Improved
increase channel sparsity from Sc=0 to Sc=0.4 (ablation)Chase · ImageNet, ResNet-50, 80% sparsityaccuracy: 75.87 -> 75.27Hurt
also prune skip-connection layers (ablation)Chase · ImageNet, ResNet-50, 90% sparsity, Sc=0.5accuracy / throughput: 73.39 -> 72.71; 78.19 -> 99.97Mixed

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