AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Prune&Comp: Free Lunch for Layer-Pruned LLMs via Iterative Pruning with Magnitude Compensation

Xinrui Chen, Hongxing Zhang, Fanyi Zeng, Yongxian Wei, Yizhi Wang, Xitong Ling, Guanghao Li, Chun Yuan

AAAI 2026 · Architectures and efficiency › Pruning and efficient tuning · approach: Layer-wise pruning of LLMs · 7 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 iterative pruning to one-shot pruning (ablation)CosSim(BI), LLaMA-3-8B 7/32 · PPL averageperplexity: 58.43 -> 38.37Improved
add magnitude compensation to one-shot pruning (ablation)CosSim(BI), LLaMA-3-8B 7/32 · PPL averageperplexity: 58.43 -> 37.39Improved
add magnitude compensation to iterative pruning (ablation)CosSim(BI), LLaMA-3-8B 7/32 · PPL averageperplexity: 38.37 -> 27.16Improved
add Prune&Comp to Taylor+LLaMA-3-8B 5/32 · PPL averageperplexity: 512.78 -> 16.34Improved
add Prune&Comp to Taylor+LLaMA-3-8B 7/32 · PPL averageperplexity: 2840.38 -> 25.21Improved
add Prune&Comp to CosSim(CL)LLaMA-3-8B 7/32 · PPL averageperplexity: 2839.30 -> 230.73Improved
add Prune&Comp to Mag+LLaMA-2-7B 7/32 · PPL averageperplexity: 89.61 -> 28.20Improved
add Prune&Comp to Taylor+LLaMA-3-8B 5/32 · QA relative performancerelative performance: 65.85 -> 90.57Improved
add Prune&Comp to CosSim(BI)LLaMA-3-8B 5/32 · QA relative performancerelative performance: 89.18 -> 93.19Improved
add Prune&Comp to CosSim(CL)Qwen3-8B 5/36 · QA relative performancerelative performance: 84.42 -> 89.29Improved
add Prune&Comp to PPL metricQwen3-8B 5/36 · QA relative performancerelative performance: 91.98 -> 89.22Hurt
add Prune&Comp to Taylor+LLaMA-3-8B 5/32 · MMLUweighted accuracy: 38.64 -> 48.42Improved

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