AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Beyond Low-Rank Tuning: Model Prior-Guided Rank Allocation for Effective Transfer in Low-Data and Large-Gap Regimes.

Chuyan Zhang, Kefan Wang, Yun Gu

ICCV 2025 · 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
allocate rank per layer by stable rank instead of uniform rank 256LoRA · VTAB-Specialized 1-shotaccuracy (%): 54.35 -> 56.38Improved
allocate rank by stable rank instead of uniform rank 64 (similar trainable parameters) (ablation)LoRA · VTAB-Specialized 1-shotaccuracy (%): 48.31 -> 57.69Improved
raise LoRA rank from 8 to 256LoRA · VTAB-Specialized 1-shotaccuracy (%): 42.99 -> 54.35Improved
add stochastic partial updating (SPU) to fixed-rank LoRA r256 (ablation)LoRA-r256 · VTAB-Specialized 1-shotaccuracy (%): 54.35 -> 53.69No clear effect
add stochastic partial updating to fixed rank 8 (ablation)LoRA-r8 · MedFM 10-shotAUC (%): 74.84 -> 78.82Improved
add stochastic partial updating to SR-LoRA (ablation)SR-LoRA · MedFM 1-shotAUC (%): 71.32 -> 72.47Improved
add stochastic partial updating to SR-LoRA (ablation)SR-LoRA · VTAB-Specialized 1-shotaccuracy (%): 57.69 -> 56.38Hurt
replace full fine-tuning with SR-LoRA on ViT-LViT-L (ImageNet-21k) · VTAB-Specialized 1-shotaccuracy (%): 46.26 -> 53.37Improved
replace LoRA with SR-LoRA on DINO ViT-LDINO ViT-L · VTAB-Specialized 1-shotaccuracy (%): 36.44 -> 43.59Improved
replace LoRA with SR-LoRA on MAE ViT-LMAE ViT-L · VTAB-Specialized 1-shotaccuracy (%): 32.80 -> 46.10Improved
replace full fine-tuning with SR-LoRAViT-B/16 · MedFM 1-shotAUC (%): 67.31 -> 72.47Improved

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