AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Sketch-Based Low-Rank Model Merging with Shared Circulant Transforms

Zhiming Zhang, Rong Yin, Xiaoshuai Hao, Hansong Zhang, Hao Peng, Yong Liu, Can Ma, Dan Meng

ICML 2026 · Architectures and efficiency › Pruning and efficient tuning · approach: LoRA variants and theory

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 shared circulant transform before sketching (ablation)CircuMerge · DARE-TIES, ViT-B/32 vision mergeavg normalized accuracy: 57.92 -> 69.77Improved
use circulant instead of dense random invertible transform (ablation)CircuMerge · DARE-TIES, ViT-B/32 vision mergeavg normalized accuracy: 62.49 -> 69.77Improved
use circulant instead of dense random invertible transform (ablation)CircuMerge · DARE-TIES, ViT-B/32 vision mergemerge time (s): 5.48 -> 3.72Improved
add random invertible shared transform vs none (ablation)CircuMerge · DARE-TIES, ViT-B/32 vision mergeavg normalized accuracy: 57.92 -> 62.49Improved
increase sketch size s beyond 8 (ablation)CircuMerge · NLI and vision suitesavg normalized accuracyHurt
merge in sketch space with CircuMerge instead of core space (DARE-TIES)DARE-TIES · 6-task NLI (Llama 3 8B LoRA merge)avg normalized accuracy: 93.32 -> 93.64Improved
merge in sketch space with CircuMerge instead of core space (TSV)TSV · 6-task NLI (Llama 3 8B LoRA merge)avg normalized accuracy: 94.16 -> 94.13No clear effect
merge in sketch space with CircuMerge instead of core space (Iso-C)Iso-C · 6-task NLI (Llama 3 8B LoRA merge)avg normalized accuracy: 90.33 -> 89.37Hurt
merge in sketch space with CircuMerge instead of core space (TSV+Iso-C)TSV+Iso-C · 8-task vision (ViT-B/32 LoRA merge)avg normalized accuracy: 76.43 -> 77.53Improved
merge in sketch space with CircuMerge instead of core space (TIES)TIES · 8-task vision (ViT-B/32 LoRA merge)avg normalized accuracy: 68.63 -> 68.39No clear effect
CircuMerge instead of KnOTS merge timeTIES · 6-task NLI (Llama 3 8B LoRA merge)merge time (s): 2963 -> 3.7Improved

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