AI Paper Atlas

Research map / Label-efficient and robust learning / Test-time adaptation

IMSE: Intrinsic Mixture of Spectral Experts Fine-tuning for Test-Time Adaptation

Sunghyun Baek, Jaemyung Yu, Seunghee Koh, Minsu Kim, Hyeonseong Jeon, Junmo Kim

ICLR 2026 · Label-efficient and robust learning › Test-time adaptation · approach: Realistic and stable test-time adaptation · 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
Add diversity maximization loss to entropy minimization (ablation)IMSE (TTA) · ImageNet-C 50k, ViT-Baccuracy (%): 67.8 -> 69.1Improved
Use diversity loss alone without entropy minimization (ablation)IMSE (TTA) · ImageNet-C 50k, ViT-Baccuracy (%): 69.1 -> 32.7Hurt
Add diversity maximization loss under non-iid label imbalance (ablation)IMSE · ImageNet-C non-iid, alpha=0.05accuracy (%): 67.33 -> 68.41Improved
Add diversity maximization loss (ablation)IMSE-Retrieval (CTTA) · ImageNet-C 5k CTTAaccuracy (%): 59.4 -> 62.2Improved
Add domain bank retrieval (ablation)IMSE-Retrieval (CTTA) · ImageNet-C 5k CTTAaccuracy (%): 59.4 -> 62.8Improved
Combine diversity loss and domain bank (ablation)IMSE-Retrieval (CTTA) · ImageNet-C 5k CTTAaccuracy (%): 62.2 -> 64.4Improved
Use mean+variance descriptor instead of mean or variance alone (ablation)IMSE-Retrieval · ImageNet-C 5k CTTAaccuracy (%): 63.1 / 63.2 -> 64.4Improved
Tune top-R% instead of bottom-R% singular values (R=20) (ablation)IMSE · ImageNet-C TTAaccuracy (%): 64.7 -> 67.7Improved
Tune top vs bottom 80% singular values (ablation)IMSE · ImageNet-C TTAaccuracy (%): 68.4 -> 69.1Improved
Adapt all linear layers instead of q,k,v,proj,fc2 (ablation)IMSE · ImageNet-C TTAaccuracy (%): 69.0 -> 69.2Improved
Raise domain-change threshold tau from 0.02 to 0.04 (ablation)IMSE-Retrieval · ImageNet-C Gradual CTTAaccuracy (%): 74.9 -> 75.6 (detection F1 0.77 -> 0.28)Mixed
Scale diversity loss weight lambda_dm from 50 up to 200 (ablation)IMSE · ImageNet-C TTAaccuracy (%): 69.0 -> ~68No clear effect

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