AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Label-Free Cross-Task LoRA Merging with Null-Space Compression

Wonyoung Lee, Wooseong Jeong, Kuk-Jin Yoon

CVPR 2026 · Architectures and efficiency › Pruning and efficient tuning · approach: LoRA variants and theory · 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 NSC null-space objective instead of entropy surrogate (AdaMerging) for NLI mergingLoRA merging · 6 NLI tasks (LLaMA-3 8B)avg normalized accuracy (%): 89.6 -> 92.3Improved
use NSC instead of gradient-free merging (TA/TIES/DARE-TIES) on dense vision tasksViT-B LoRA merging · 20 vision tasksavg normalized performance (%): 77.2 -> 92.0 (TA vs NSC)Improved
NSC instead of AdaMerging Single Token on VLM mergingLLaVA-1.5-7B LoRA merging · 6 VLM tasksavg normalized score (%): 80.9 -> 82.7Improved
NSC instead of AdaMerging Full Token on VLM mergingLLaVA-1.5-7B LoRA merging · 6 VLM tasksavg normalized score (%): 82.4 -> 82.7No clear effect
NSC instead of AdaMerging Full Token for optimization costLLaVA-1.5-7B LoRA merging · LLaVA 6 taskstotal time (min): 104.0 -> 13.3Improved
NSC instead of gradient-free merging on unseen tasksViT-B LoRA merging · 20 vision tasks, unseen splitavg normalized performance (%): 76.8 -> 87.1 (TA vs NSC)Improved
restrict NSC objective to only 1 transformer block instead of 3-6 (ablation)NSC objective (O projection) · 8 image classification tasks (CLIP ViT-B/32)accuracy (%): 84.6 -> 82.6Hurt
activate all 12 blocks instead of 3-6 for the NSC objective (ablation)NSC objective (O projection) · 8 image classification tasks (CLIP ViT-B/32)accuracy (%): 84.9 (6 blocks) -> 84.5 (12 blocks)No clear effect
target only O projection instead of QKVO (ablation)NSC objective · 8 image classification tasks (CLIP ViT-B/32)accuracy (%): 83.7 -> 84.5 (12 blocks)No clear effect
gradient-free low-rank SVD or Linear merging instead of NSCViT-B LoRA merging · 20 vision tasksavg normalized performance (%): SVD 83.0, Linear 74.3 vs NSC 92.0Mixed

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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