AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Efficient Orthogonal Fine-Tuning with Principal Subspace Adaptation

Fei Wu, Jia Hu, Geyong Min, Shiqiang Wang

ICLR 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
confine orthogonal fine-tuning to the principal subspace (PSOFT) instead of full-space GOFTv2PSOFT · DeBERTaV3-base GLUEpeak memory (GB): 18.5 -> 4.1Improved
use PSOFT instead of LoRALLaMA-3.2-3B · GSM-8Kaccuracy: 60.80 -> 63.08Improved
use PSOFT instead of FFT on MATHLLaMA-3.2-3B · MATHaccuracy: 16.84 -> 15.98Hurt
use PSOFT instead of LoRALLaMA-3.1-8B · Commonsense reasoning avgaverage accuracy: 81.11 -> 82.54Improved
use PSOFT instead of LoRA on ViT-B/16ViT-B/16 · VTAB-1Kaverage accuracy: 71.8 -> 73.4Improved
orthogonal initialization of A (A_orth R B) instead of PiSSA initialization (ablation)PSOFT · RTE (DeBERTaV3-base)accuracy: 71.11 -> 85.92Improved
enforce orthogonality on B instead of A at initialization (ablation)PSOFT · RTE (DeBERTaV3-base)accuracy: 85.92 -> 52.71Hurt
enable both tunable vectors alpha and beta (ablation)PSOFT · LLaMA-3.2-3B rank 64, GSM-8K and MATHaccuracyImproved
strict orthogonality via Cayley parameterization versus unconstrained R (gamma=0) (ablation)PSOFT · ablation on orthogonalityaccuracyImproved
add DoRA weight decomposition to LoRALoRA · ViT-B/16 VTAB-1Kpeak memory (GB): 9.9 -> 17.8Hurt

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