AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

QR-LoRA: Efficient and Disentangled Fine-tuning via QR Decomposition for Customized Generation

Jiahui Yang, Yongjia Ma, Donglin Di, Jianxun Cui, Hao Li, Wei Chen, Yan Xie, Xun Yang, Wangmeng Zuo

ICCV 2025 · Architectures and efficiency › Pruning and efficient tuning · approach: LoRA variants and theory · 12 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
train only the delta-R matrix with fixed Q and R instead of training full LoRA A and B (ablation)QR-LoRA · content-style fusioncosine similarity between task updates: max cosine similarity between delta-R matrices below 0.2Improved
directly fine-tune Q and R instead of only delta-R (ablation)QR-LoRA · content-style fusionfeature disentanglement (cosine similarity): R matrix similarity > 0.75Hurt
replace B-LoRA / ZipLoRA with QR-LoRA on SDXLSDXL · content-style fusionuser study rating (0-5): 3.34 -> 4.07 (vs B-LoRA)Improved
replace direct merge DB-LoRA with QR-LoRA on SD3SD3 · content-style fusionuser study rating (0-5): 2.93 -> 4.14Improved
replace direct merge DB-LoRA with QR-LoRA on FLUX.1-devFLUX.1-dev · content-style fusionDINO-C: 0.654 -> 0.687Improved
vary merge scaling coefficients lambda_c and lambda_s (ablation)QR-LoRA · content-style fusiongeneration qualityNo clear effect
train only delta-R versus standard LoRA (ablation)QR-LoRA · convergenceconvergence speedNo clear effect
train both Q and R instead of only delta-R (ablation)QR-LoRA · convergenceconvergence speedImproved
use QR decomposition-based delta-R reduces trainable parametersQR-LoRA · parameter counttrainable parameters: half of conventional LoRAImproved

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