AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

ReFTA: Breaking the Weight Reconstruction Bottleneck in Tensorized Parameter-Efficient Fine-Tuning

Jingjing Zheng, Anda Tang, Qiangqiang Mao, Zhouchen Lin, Yankai Cao

CVPR 2026 · Architectures and efficiency › Pruning and efficient tuning · approach: Adapter-based parameter-efficient tuning

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
fine-tune only principal tensor components (T-SVD) instead of LoRA-style adaptersReFTA on ViT-Large · Image classification avg (5 datasets)average accuracy: 79.99 -> 85.67Improved
ReFTA vs LoRA (r = 8) with far fewer parametersViT-Huge · Image classification avg (OxfordPets, StanfordCars, FGVC)average accuracy: 75.23 -> 76.32Improved
ReFTA vs LoRA on RoBERTa-Base GLUEReFTA · GLUE avg (6 datasets)average score: 85.2 (LoRA) vs 84.8No clear effect
ReFTA vs PiSSA/LoRA-PRO/LoRETTA at small rank on RoBERTa-LargeReFTA · GLUE avg (6 datasets)average score: 83.2 -> 88.2 (vs LoRA-PRO)Improved
ReFTA (r=15) vs LoRA (r=8) on CommonsenseQA with ~94% fewer parametersReFTA · CommonsenseQAaccuracy: 86.11 -> 85.91 (avg)No clear effect
use DCT or LSM-3 invertible transform instead of identity (ablation)ReFTA · StanfordCars, FGVC (ViT-Large), CoLA, RTE (RoBERTa-Large)accuracy: up to 3-5% on vision tasksImproved
make the transform U0 trainable (init DCT) (ablation)ReFTA · FGVC (ViT-Large)accuracyNo clear effect
fine-tune only principal components (residual quantized) vs Gaussian-Zero init variant (ablation)ReFTA · Quantization error at init (ViT-Large query weights)quantization errorImproved
increase tensor rank R from 5 to 15 (ablation)ReFTA on ViT-Huge · Image classification avg (OxfordPets, StanfordCars, FGVC)average accuracy: 74.52 -> 76.32Improved
ReFTA vs LoRETTA on ViT-LargeReFTA · Image classification avg (5 datasets)average accuracy: 78.51 -> 85.67Improved

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