AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

AIRA: Activation-Informed Low-Rank Adaptation for Large Models

Lujun Li, Dezhi Li, Cheng Lin, Wei Li, Wei Xue, Sirui Han, Yike Guo

ICCV 2025 · Architectures and efficiency › Pruning and efficient tuning · approach: LoRA variants and theory · 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
add outlier-weighted SVD initialization instead of Kaiming (ablation)AIRA · few-shot 4-shottop-1 accuracy: 79.8 -> 80.2Improved
add outlier-weighted SVD initialization instead of Kaiming (ablation)AIRA · few-shot 16-shottop-1 accuracy: 83.6 -> 84.6Improved
add outlier-weighted SVD initialization instead of standard SVD (ablation)AIRA · few-shot 4-shottop-1 accuracy: 79.9 -> 80.2Improved
add outlier-weighted SVD initialization instead of standard SVD (ablation)AIRA · few-shot 16-shottop-1 accuracy: 84.3 -> 84.6Improved
add outlier-driven rank assignment instead of uniform (ablation)AIRA · few-shot 4-shottop-1 accuracy: 78.2 -> 80.2Improved
add outlier-driven rank assignment instead of uniform (ablation)AIRA · few-shot 16-shottop-1 accuracy: 83.8 -> 84.6Improved
add outlier-driven rank assignment instead of growth (ablation)AIRA · few-shot 16-shottop-1 accuracy: 84.0 -> 84.6Improved
add input-weighted normalization instead of original LoRA training (ablation)AIRA · few-shot 4-shottop-1 accuracy: 79.0 -> 80.2Improved
add input-weighted normalization instead of output-weighted (ablation)AIRA · few-shot 16-shottop-1 accuracy: 84.0 -> 84.6Improved
use -x objective instead of -log(x) (ablation)AIRA · few-shot 4-shottop-1 accuracy: 65.5 -> 64.4Hurt
use learning rate 1e-3 instead of 1e-4 (ablation)AIRA · few-shot 16-shottop-1 accuracy: 70.8 -> 73.4Improved
use AIRA instead of PiSSA on Mistral-7BAIRA · GSM8K+MATHtop-1 accuracy: 44.35 -> 45.23Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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