AI Paper Atlas

Research map / Vision-language and multimodal / CLIP and zero-shot

Robustifying Vision-Language Models via Test-Time Prompt Adaptation

Xingyu Zhu, Huanshen Wu, Shuo Wang, Beier Zhu, Jiannan Ge, Jiaheng Zhang, Long Chen

ICML 2026 · Vision-language and multimodal › CLIP and zero-shot · approach: Test-time adaptation of VLMs · 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
use RITA optimal-transport distribution alignment instead of R-TPTCLIP ViT-B/16 · 8 fine-grained datasets, eps=1.0average robust accuracy: 50.5 -> 52.7Improved
use RITA instead of R-TPTCLIP ViT-B/32 · 8 fine-grained datasets, eps=1.0average robust accuracy: 47.9 -> 49.7Improved
use RITA instead of R-TPT on CarsCLIP ViT-B/32 · Cars, eps=1.0robust accuracy: 44.6 -> 42.9Hurt
plug RITA onto adversarially finetuned FARE CLIPFARE · 8 fine-grained datasets avgrobust accuracy: 23.3 -> 41.0Improved
use RITA on top of PMGPMG CLIP · 8 fine-grained datasets avgrobust accuracy: 26.1 -> 32.6Improved
use RITA against CW attackCLIP ViT-B/16 · 8 fine-grained avg, CWrobust accuracy: 51.2 -> 54.0Improved
use RITA against DI attackCLIP ViT-B/16 · 8 fine-grained avg, DIrobust accuracy: 44.2 -> 46.2Improved
add dynamic cache term (alpha 0 to 0.1) (ablation)RITA · DTD and ImageNet, ViT-B/16accuracyImproved
increase number of learnable prompts from 4 to 32 (ablation)RITA · DTD, ViT-B/16clean and robust accuracyImproved
increase augmented views from 96 to 128 (ablation)RITA · DTD, ViT-B/16robust accuracyHurt
increase TTA steps from 1 to 2 (ablation)RITA · fine-grained datasets and ImageNetrobust accuracy: +0.5 to +0.6Improved
increase TTA steps from 2 to 4 (ablation)RITA · fine-grained datasets and ImageNetrobust accuracyHurt

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