AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

DWTSG: Parameter-Efficient Fine-Tuning of Large Pre-trained Models via Discrete Wavelet Transform and Subband Guidance

Chengwei Sun, Jiwei Wei, Shiyuan He, Zeyu Ma, Yuyang Zhou, Ran Ran, Jie Zou, Yang Yang

AAAI 2026 · Architectures and efficiency › Pruning and efficient tuning · approach: Adapter-based parameter-efficient tuning · 0 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 DWTSG instead of LoRA with fewer parametersRoBERTa-large · GLUE avgaverage score: 88.6 -> 88.8Improved
use DWTSG instead of FourierFTRoBERTa-large · GLUE avgaverage score: 88.2 -> 88.8Improved
use DWTSG instead of full fine-tuningRoBERTa-large · GLUE avgaverage score: 88.9 -> 88.8Hurt
use DWTSG instead of MiLoRALLaMA2-7B · Commonsense170K avgaverage accuracy: 79.2 -> 80.0Improved
use DWTSG instead of LoRAViT-base · image classification avgaccuracy (%): 77.6 -> 82.4Improved
use DWTSG instead of full fine-tuningViT-large · image classification avgaccuracy (%): 90.2 -> 89.6Hurt
use sym4 wavelet instead of haar, db4 or coif1 (ablation)DWTSG · ViT-base, four image datasetsaccuracy (%): best on all four, up to 5.3 points over other basesImproved
use role-aware subband allocation instead of Random (ablation)DWTSG · ViT-base OxfordPetsaccuracy (%): 95.6 -> 96.7Improved
use role-aware subband allocation instead of Energy (ablation)DWTSG · ViT-base FGVCaccuracy (%): 38.5 -> 39.7Improved
use Energy allocation instead of Random (ablation)DWTSG · ViT-base OxfordPetsaccuracy (%): 95.6 -> 94.6Hurt
raise scaling factor alpha from 0.3 to 1.1 (ablation)DWTSG · ViT-base FGVCaccuracy (%): 39.7 -> 5.1Hurt
raise scaling factor alpha from 0.1 to 0.9 (ablation)DWTSG · ViT-base OxfordPets and StanfordCarsaccuracy (%): 93.1 -> 96.7 (OxfordPets)Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 2 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