AI Paper Atlas

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

Bilateral Information-aware Test-time Adaptation for Vision-Language Models

Jingwei Sun, Jianing Zhu, Jiangchao Yao, Gang Niu, Masashi Sugiyama, Bo Han

ICLR 2026 · Label-efficient and robust learning › Test-time adaptation · approach: Pseudo-label test-time adaptation

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 BITTA high-entropy unlearning to BATBAT, CLIP ViT-B-16 · CIFAR-10-Cmean accuracy: 73.34 -> 74.78Improved
add BITTA to BATBAT, CLIP ViT-B-32 · CIFAR-100-Cmean accuracy: 36.29 -> 38.23Improved
add BITTA to BATBAT, CLIP ViT-B-16 · ImageNet-Cmean accuracy: 31.06 -> 31.88Improved
add BITTA to TPTTPT, ViT-B-16 · CIFAR-10-Cmean accuracy: 63.84 -> 65.37Improved
add BITTA to TPSTPS, ViT-B-16 · CIFAR-10-Cmean accuracy: 64.28 -> 65.24Improved
add BITTA to CTPTCTPT, ViT-B-16 · CIFAR-10-Cmean accuracy: 61.84 -> 62.61Improved
add BITTA to SARSAR, ViT-Base · ImageNet-C Gaussianaccuracy: 52.83 -> 56.27Improved
add BITTA to BATBAT, ViT-B-16 · CIFAR-10-CECE: 15.25 -> 12.93Improved
unlearning high-entropy samples instead of weight decay regularization (ablation)BAT · CIFAR-10-Cmean accuracy: 73.79 -> 74.78Improved
unlearning instead of distribution dissolve regularization (ablation)BAT · CIFAR-10-Cmean accuracy: 74.37 -> 74.78Improved
BITTA effect on individual corruption typesBAT, ViT-B-16 · ImageNet-C Frost/Brightness/Elasticaccuracy: -1.02, -0.64, -1.22Mixed
use a different prompt template (ablation)BITTA · CIFAR-10-C ViT-B-16accuracy: 74.78 -> 75.55No 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