AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Zeroth-Order Fine-Tuning of LLMs in Random Subspaces

Ziming Yu, Pan Zhou, Sike Wang, Jia Li, Mi Tian, Hua Huang

ICCV 2025 · Architectures and efficiency › Pruning and efficient tuning · approach: Adapter-based parameter-efficient tuning · 15 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
add column-orthogonal projection matrices (ablation)SubZero · RTE (OPT-13B, FT)accuracy: 67.5 -> 74.0Improved
add column-orthogonal projection matrices (ablation)SubZero · WSC (OPT-13B, FT)accuracy: 59.6 -> 65.1Improved
add reshaping of non-square matrices (ablation)SubZero · SST-2 prompt tuning (OPT-1.3B)accuracy: 74.2 -> 89.1Improved
add reshaping of non-square matrices (ablation)SubZero · SST-2 prefix tuning (OPT-1.3B)accuracy: 89.4 -> 92.2Improved
add reshaping of non-square matrices (ablation)SubZero · SST-2 LoRA (OPT-1.3B)accuracy: 92.1 -> 92.9Improved
use random projection matrix instead of activation matrix (ablation)SubZero · RTE (OPT-13B, FT)accuracy: 53.1 -> 74.0Improved
use random projection matrix instead of weight matrix (ablation)SubZero · RTE (OPT-13B, FT)accuracy: 65.3 -> 74.0Improved
update subspace every 1000 steps instead of 20000 (rank 128) (ablation)SubZero · RTE (OPT-13B, FT)accuracy: 68.6 -> 74.0Improved
vary subspace rank r among 32, 64, 128 (ablation)SubZero · RTE (OPT-13B, FT)accuracy: 72.6 / 70.0 / 72.2No clear effect
replace MeZO with SubZero in prompt tuningMeZO · CB prompt tuning (LLaMA2-7B)accuracy: 60.7 -> 66.1Improved
replace MeZO with SubZero in full-parameter tuningMeZO · CB FT (LLaMA2-7B)accuracy: 64.3 -> 71.4Improved
replace S-RGF subspace method with SubZero for memoryS-RGF · SST-2 FT memory (RoBERTa-large)memory (GB): 23.845 -> 2.690Improved

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