AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Movement Pruning: Adaptive Sparsity by Fine-Tuning

Victor Sanh, Thomas Wolf, Alexander Rush

NeurIPS 2020 · Architectures and efficiency › Pruning and efficient tuning · approach: Structured network pruning · 33 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
replace magnitude pruning with movement pruningBERT-base · SQuAD dev, 3% weightsF1: 54.5 -> 76.3Improved
use soft movement pruning (threshold + regularizer) instead of hard Top-v movement pruning (ablation)BERT-base · SQuAD dev, 3% weightsF1: 76.3 -> 79.9Improved
replace L0 regularization with movement pruningBERT-base · SQuAD dev, 3% weightsF1: 73.3 -> 76.3Improved
use movement pruning instead of magnitude pruning at low sparsity (>70% remaining weights)BERT-base · MNLI/SQuAD/QQPaccuracyHurt
add knowledge distillation from dense BERT-base teacher (ablation)movement pruning · SQuAD dev, 10% weightsF1: 81.7 -> 84.3Improved
add knowledge distillation (ablation)soft movement pruning · SQuAD dev, 3% weightsF1: 79.9 -> 82.3Improved
add knowledge distillation (ablation)soft movement pruning · QQP dev, 3% weightsaccuracy: 89.3 -> 89.1No clear effect
use global instead of local Top-v mask selection (ablation)movement pruning · SQuAD, 3% weightsF1: +2.3 F1Mixed

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