AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Sparsity Outperforms Low-Rank Projections in Few-Shot Adaptation

Nairouz Mrabah, Nicolas Richet, Ismail Ben Ayed, Eric Granger

ICCV 2025 · Architectures and efficiency › Pruning and efficient tuning · approach: Efficient ViT adaptation · 0 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
sparsify the gradient (random Top-M selection) instead of dense Adam (ablation)SO · 9 datasets 1-shotaverage top-1 accuracy: 4.4 -> 73.6Improved
add importance-pruned sparse moments (ablation)SO · 9 datasets 1-shotaverage top-1 accuracy: 73.6 -> 74.6Improved
add importance-pruned sparse moments (ablation)SO · 9 datasets 4-shotaverage top-1 accuracy: 77.2 -> 80.4Improved
use random instead of importance-based gradient selection (ablation)SO · 9 datasets 1-shotaverage top-1 accuracy: 66.6 -> 74.6Improved
use random instead of importance-based gradient selection (ablation)SO · 9 datasets 2-shotaverage top-1 accuracy: 68.3 -> 77.7Improved
use importance-based instead of random moment pruning (ablation)SO · 9 datasets 1-shotaverage top-1 accuracy: 72.9 -> 74.6Improved
use importance-based instead of random moment pruning (ablation)SO · 9 datasets 4-shotaverage top-1 accuracy: 79.8 -> 80.4Improved
use dynamic instead of static sparsity support (ablation)SO · 9 datasets 1-shotaverage top-1 accuracy: 71.2 -> 74.6Improved
use dynamic instead of static sparsity support (ablation)SO · 9 datasets 4-shotaverage top-1 accuracy: 77.0 -> 80.4Improved
add moments (ablation)SO · 9 datasets 1-shotaverage top-1 accuracy: 67.8 -> 74.6Improved
add moments (ablation)SO · 9 datasets 2-shotaverage top-1 accuracy: 73.0 -> 77.7Improved
use SO instead of ReLoRASO · 11 datasets 1-shotaverage top-1 accuracy: 72.5 -> 73.8Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 5 of 5 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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