AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Stay Unique, Stay Efficient: Preserving Model Personality in Multi-Task Merging

Kuangpu Guo, Aijing Yu, Jian Liang, Yuhe Ding, Zilei Wang, Ran He, Tieniu Tan

ECCV 2026 · Architectures and efficiency › Pruning and efficient tuning · approach: Parameter-efficient transfer analysis · 1 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
personalize merging with DTS-D instead of basic merging (Task-Arithmetic)CLIP ViT-B/32 · 8 vision tasksaccuracy: 69.22 -> 90.40Improved
use DTS-D instead of T-Switch at lower memoryCLIP ViT-B/32 · 8 vision tasksaccuracy: 90.15 -> 90.40Improved
add SVD decomposition before thresholding (ablation)DTS-T · 8 vision tasksmemory rate (accuracy nearly equal): AMR 9.37 -> 3.68; accuracy 90.41 -> 90.32Improved
add per-group scaling factors (ablation)DTS-T · 8 vision tasksaccuracy: 4.98 -> 90.32Improved
use 4-group thresholding instead of binarization (ablation)DTS · 8 vision tasksaccuracy: approximately +2%Improved
increase thresholding groups to eight (ablation)DTS · 8 vision tasksaccuracy: storage doubled, virtually no gainNo clear effect
replace median with K-Means dynamic thresholding (ablation)DTS-T · ViT-B/32 seen and unseen tasksaccuracy: seen 87.94 -> 88.25; unseen 55.23 -> 55.76 (DTS-D unseen 55.41 -> 55.37)Mixed
use semantic-similarity weighted fusion of task information for unseen tasksDTS-D · RESISC45+SVHN unseen, ViT-B/32accuracy: T-Switch 42.05 -> 55.41Improved
use DTS reconstruction instead of rank-1 LoRACLIP ViT-B/32 · 8 vision tasksaccuracy: 68.61 -> 90.32 (switching 1.01 -> 1.06 ms)Improved
cap DTS storage below 1% (DTS-D*) instead of r=0.3 (ablation)DTS-D · 8 vision tasks, ViT-B/32accuracy: 90.40 -> 90.17No clear effect
apply personalized merging on unseen tasks without adaptive weighting (baselines)T-Switch · unseen tasks, ViT-B/32accuracy: 42.05 vs Ties-Merging 53.85Hurt

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