AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Lookahead: A Far-sighted Alternative of Magnitude-based Pruning

Sejun Park*, Jaeho Lee*, Sangwoo Mo, Jinwoo Shin

ICLR 2020 · Architectures and efficiency › Pruning and efficient tuning · approach: Structured network pruning · 16 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-based scores with adjacent-layer lookahead distortion scoresFCN · MNISTtest error rate at 0.12% weights remaining: 67.62 -> 16.45Improved
Replace magnitude-based scores with adjacent-layer lookahead distortion scoresConv-6 · CIFAR-10test error rate at 3.62% weights remaining: 28.60 -> 24.52Improved
Replace magnitude-based scores with adjacent-layer lookahead distortion scoresVGG-19 · CIFAR-10test error rate at 1.24% weights remaining: 77.90{ $\pm$0.12} -> 22.82{ $\pm$0.81}Improved
Replace magnitude-based scores with adjacent-layer lookahead distortion scoresResNet-18 · CIFAR-10test error rate at 0.36% weights remaining: 26.77{ $\pm$1.04} -> 22.45{ $\pm$0.64}Improved
Replace magnitude-based scores with adjacent-layer lookahead distortion scoresResNet-50 · Tiny-ImageNettop-5 test error rate at 0.72% weights remaining: 59.18 $\pm16.31$ -> 46.16 $\pm1.04$Improved
Use only succeeding-layer lookahead (ablation)LAP · MNISTtest error rate at 0.12% weights remaining: 16.45 -> 50.11Hurt
Use only preceding-layer lookahead (ablation)LAP · MNISTtest error rate at 0.12% weights remaining: 16.45 -> 28.51Hurt
Prune layers in forward order using already-pruned preceding layers (ablation)LAP · MNISTtest error rate at 0.12% weights remaining: 16.45 -> 12.00Improved
Prune layers in backward order using already-pruned succeeding layers (ablation)LAP · MNISTtest error rate at 0.12% weights remaining: 16.45 -> 13.00Improved
Apply forward-order pruning gradually over five steps (ablation)LAP · MNISTtest error rate at 0.12% weights remaining: 16.45 -> 10.98Improved
Apply backward-order pruning gradually over five steps (ablation)LAP · MNISTtest error rate at 0.12% weights remaining: 16.45 -> 11.92Improved
Use estimated activation probabilities to rescale lookahead scores (ablation)LAP · MNISTtest error rate at 0.12% weights remaining: 11.01 -> 7.99Improved

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