AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Effective Model Pruning : Measuring the Redundancy of Model Components

Yixuan Wang, Dan P. Guralnik, Saiedeh Akbari, Warren Dixon

ICML 2026 · Architectures and efficiency › Pruning and efficient tuning · approach: Structured network pruning

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
choose sparsity adaptively with EMP instead of fixed 50% for magnitude pruning (ablation)magnitude pruning on LLaMA/LLaMA-2 · WikiText avg over 7 modelsavg change in perplexity: +2.982 -> +0.752 at sparsity 50.00 -> 36.63Improved
choose sparsity adaptively with EMP instead of fixed 50% for magnitude pruning (ablation)magnitude pruning on LLaMA/LLaMA-2 · zero-shot 7 tasks avgavg change in accuracy (%): -2.60 -> -0.93Improved
apply EMP threshold to Wanda (ablation)Wanda · LLaMA/LLaMA-2 avgavg change in perplexity: +0.799 -> +0.678 at sparsity 50.00 -> 40.47No clear effect
EMP-Wanda on LLaMA-2 7B (ablation)Wanda · LLaMA-2 7B WikiTextperplexity: 6.410 -> 6.513Hurt
prune with beta below 1 (beyond n_eff) (ablation)EMP on FC models · MNIST / Fashion-MNISTaccuracyHurt
keep more than n_eff entries (beta above 1) (ablation)EMP on FC models · MNIST / Fashion-MNISTaccuracyNo clear effect
apply EMP threshold to magnitude pruning of VGG16 (ablation)VGG16 CIFAR-10 · CIFAR-10accuracy (%): 69.08 (90% fixed) -> 90.98 (EMP, 59.5% sparsity)Improved
apply EMP to Taylor loss-change criterion (ablation)VGG16 CIFAR-10 · CIFAR-10sparsity (%): accuracy 91.12 maintained at sparsity 2.1Mixed
EMP structured pruning at beta=1 on F-MNIST FC5FC5 · F-MNISTaccuracy (pp): 84.57 -> 86.44Improved
EMP structured pruning at beta=1 on CIFAR-10 VGG16VGG16 · CIFAR10accuracy (pp): 91.12 -> 90.69No clear effect
apply EMP to KAN node pruningKAN [784,64,10] · MNISTvalidation accuracy (%): 97.15 -> 94.36Hurt
EMP head pruning at beta=1 with Taylor criterionGPT-2 · WikiText-2perplexity: 34.86 -> 35.20No 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