AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

GAPrune: Gradient-Alignment Pruning for Domain-Aware Embeddings

Yixuan Tang, Yi Yang

ICLR 2026 · Architectures and efficiency › Pruning and efficient tuning · approach: Layer-wise pruning of LLMs · 2 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
use GAPrune (Fisher plus gradient alignment) instead of Domain Fisher pruning (ablation)Qwen3-Embedding-4B, 50% sparsity · FinMTEB one-shotaverage score: 0.4887 -> 0.5224Improved
use GAPrune instead of General Fisher pruning (ablation)Qwen3-Embedding-4B, 50% sparsity · FinMTEB one-shotaverage score: 0.3623 -> 0.5224Improved
use GAPrune instead of magnitude pruning (ablation)Qwen3-Embedding-4B, 50% sparsity · ChemTEB one-shotaverage score: 0.7299 -> 0.7462Improved
use GAPrune instead of magnitude pruning (ablation)Qwen3-Embedding-4B, 30% sparsity · FinMTEB one-shotaverage score: 0.5415 -> 0.5425No clear effect
use GAPrune instead of magnitude pruning (ablation)E5-Mistral-7B-Instruct, 50% sparsity · ChemTEB one-shotaverage score: 0.6189 -> 0.6206No clear effect
retrain pruned model for 100 steps after GAPruneQwen3-Embedding-4B, 50% sparsity · FinMTEBaverage score: 0.5224 -> 0.5594Improved
retrain pruned model for 100 steps after GAPruneQwen3-Embedding-4B, 50% sparsity · ChemTEBaverage score: 0.7462 -> 0.7770Improved
use GAPrune instead of L3 Prune after retraining (ablation)E5-Mistral-7B-Instruct · ChemTEB prune-and-retrainaverage score: 0.6687 -> 0.6773Improved
use GAPrune instead of Domain Fisher pruning for embedding geometry (ablation)Qwen3-Embedding-4B · alignment loss, effective dimensionalignment loss: 0.75 -> 0.51Improved
use GAPrune instead of magnitude pruning for uniformity (ablation)Qwen3-Embedding-4B · embedding uniformity lossuniformity loss: -3.30 -> -2.41Hurt
random pruning of parameters at 50% sparsity (ablation)Qwen3-Embedding-4B · FinMTEB one-shotaverage score: 0.5353 -> 0.2165Hurt

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