AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Generalized Tensor-based Parameter-Efficient Fine-Tuning via Lie Group Transformations

Chongjie Si, Zhiyi Shi, Xuehui Wang, Yichen Xiao, Xiaokang Yang, Wei Shen

ICCV 2025 · Architectures and efficiency › Pruning and efficient tuning · approach: Adapter-based parameter-efficient tuning · 3 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
replace additive LoRA update with Lie group multiplicative update (LieRA)LoRA r=16 ConvNeXt-V2-B · VTAB-1k averageaccuracy: 74.1 -> 75.5Improved
replace additive LoRA update with LieRALoRA r=2 ConvNeXt-V2-B · VTAB-1k averageaccuracy: 74.4 -> 74.7Improved
replace additive LoRA update with LieRALoRA r=16 ConvNeXt-V2-B · MS COCO detectionmAP: 35.5 -> 39.1Improved
replace additive LoRA update with LieRALoRA r=32 ConvNeXt-V2-B · MS COCO detectionmAP: 35.9 -> 40.5Improved
replace additive LoRA update with LieRALoRA r=16 LLaMA-7B · Commonsense reasoning averageaccuracy: 70.9 -> 75.2Improved
replace additive LoRA update with LieRALoRA r=32 LLaMA3-8B · Commonsense reasoning averageaccuracy: 80.8 -> 85.3Improved
replace additive LoRA update with LieRALoRA r=2 DeBERTaV3-base · GLUE dev averageaverage score: 88.13 -> 88.97Improved
replace additive LoRA update with LieRALoRA r=8 DeBERTaV3-large · GLUE dev averageaverage score: 90.79 -> 90.98Mixed
replace additive LoRA update with LieRA on LLaMA-7BLoRA r=16 LLaMA-7B · BoolQaccuracy: 69.9 -> 68.5Hurt
add first-order Taylor approximation to LieRA (ablation)LieRA r=16 · VTAB-1k averageaccuracy: 75.7 -> 75.5No clear effect
add first-order Taylor approximation to LieRA (ablation)LieRA r=16 · MS COCO training costtime per epoch (min): 76.28 -> 50.17Improved
add first-order Taylor approximation to LieRA (ablation)LieRA r=16 · MS COCO training costGPU memory (GB): 14.74 -> 9.97Improved

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