{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,4]],"date-time":"2026-05-04T03:28:26Z","timestamp":1777865306607,"version":"3.51.4"},"reference-count":63,"publisher":"IEEE","license":[{"start":{"date-parts":[[2025,10,19]],"date-time":"2025-10-19T00:00:00Z","timestamp":1760832000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,10,19]],"date-time":"2025-10-19T00:00:00Z","timestamp":1760832000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,10,19]]},"DOI":"10.1109\/iccv51701.2025.00280","type":"proceedings-article","created":{"date-parts":[[2026,4,29]],"date-time":"2026-04-29T19:45:49Z","timestamp":1777491949000},"page":"2919-2929","source":"Crossref","is-referenced-by-count":0,"title":["Think Twice: Test-Time Reasoning for Robust CLIP Zero-Shot Classification"],"prefix":"10.1109","author":[{"given":"Shenyu","family":"Lu","sequence":"first","affiliation":[{"name":"Elmore Family School of Electrical and Computer Engineering, Purdue University"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Zhaoying","family":"Pan","sequence":"additional","affiliation":[{"name":"Elmore Family School of Electrical and Computer Engineering, Purdue University"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Xiaoqian","family":"Wang","sequence":"additional","affiliation":[{"name":"Elmore Family School of Electrical and Computer Engineering, Purdue University"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Zero-shot robustification of zero-shot models","volume-title":"The Twelfth International Conference on Learning Representations","author":"Adila"},{"key":"ref2","article-title":"Perceptionclip: Visual classification by inferring and conditioning on contexts","volume-title":"The Twelfth International Conference on Learning Representations","author":"An"},{"key":"ref3","article-title":"Invariant risk minimization","author":"Arjovsky","journal-title":"arXiv preprint"},{"key":"ref4","first-page":"3080","article-title":"Mt3: Meta test-time training for selfsupervised test-time adaption","volume-title":"International Conference on Artificial Intelligence and Statistics","author":"Bartler"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1037\/0033-295X.94.2.115"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/iccv48922.2021.00045"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.00276"},{"key":"ref8","article-title":"Debiasing visionlanguage models via biased prompts","author":"Chuang","year":"2023","journal-title":"arXiv preprint"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1109\/ISBI.2018.8363547"},{"key":"ref10","article-title":"Vision transformers need registers","volume-title":"The Twelfth International Conference on Learning Representations","author":"Darcet"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.52202\/075280-0069"},{"key":"ref12","article-title":"An image is worth 16x16 words: Transformers for image recognition at scale","volume-title":"International Conference on Learning Representations","author":"Dosovitskiy"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1038\/s42256-020-00257-z"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.52202\/079017-1998"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.01853"},{"key":"ref16","first-page":"336","article-title":"Simple data balancing achieves competitive worst-group-accuracy","volume-title":"Conference on Causal Learning and Reasoning","author":"Youbi Idrissi"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/wacv61041.2025.00532"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.02302"},{"key":"ref19","article-title":"Last layer re-training is sufficient for robustness to spurious correlations","volume-title":"The Eleventh International Conference on Learning Representations","author":"Kirichenko"},{"key":"ref20","article-title":"Last layer re-training is sufficient for robustness to spurious correlations","volume-title":"The Eleventh International Conference on Learning Representations","author":"Kirichenko"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.00371"},{"key":"ref22","article-title":"Entropy is not enough for test-time adaptation: From the perspective of disentangled factors","author":"Lee","year":"2024","journal-title":"arXiv preprint"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.01922"},{"key":"ref24","article-title":"Metashift: A dataset of datasets for evaluating contextual distribution shifts and training conflicts","volume-title":"International Conference on Learning Representations","author":"Liang"},{"key":"ref25","first-page":"6781","article-title":"Just train twice: Improving group robustness without training group information","volume-title":"International Conference on Machine Learning","author":"Liu"},{"key":"ref26","article-title":"Avoiding spurious correlations via logit correction","volume-title":"The Eleventh International Conference on Learning Representations","author":"Liu"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72970-6_3"},{"key":"ref28","first-page":"21808","article-title":"Ttt++: When does self-supervised test-time training fail or thrive?","volume":"34","author":"Liu","year":"2021","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2015.425"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1145\/3637528.3671902"},{"key":"ref31","article-title":"Mitigating spurious correlations in zero-shot multimodal models","volume-title":"The Thirteenth International Conference on Learning Representations","author":"Lu"},{"key":"ref32","article-title":"Test-time adaptation to distribution shift by confidence maximization and input transformation","author":"Kumar Mummadi","year":"2021","journal-title":"arXiv preprint"},{"key":"ref33","article-title":"Learning from failure: Training debiased classifier from biased classifier","author":"Nam","year":"2020","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref34","first-page":"16888","article-title":"Efficient test-time model adaptation without forgetting","volume-title":"International conference on machine learning","author":"Niu"},{"key":"ref35","article-title":"Dinov2: Learning robust visual features without supervision","author":"Oquab","year":"2023","journal-title":"arXiv preprint"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1201.0490"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01756"},{"key":"ref38","first-page":"8748","article-title":"Learning transferable visual models from natural language supervision","volume-title":"International conference on machine learning","author":"Radford"},{"key":"ref39","article-title":"Distributionally robust neural networks for group shifts: On the importance of regularization for worstcase generalization","author":"Sagawa","journal-title":"arXiv preprint"},{"key":"ref40","first-page":"8346","article-title":"An investigation of why overparameterization exacerbates spurious correlations","volume-title":"International Conference on Machine Learning","author":"Sagawa"},{"key":"ref41","article-title":"Zero-shot recognition with guided cropping","author":"Saranrittichai","year":"2024","journal-title":"ICLR 2024 Workshop on Mathematical and Empirical Understanding of Foundation Models"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1833"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-021-01595-0"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.5244\/c.35.365"},{"key":"ref45","first-page":"19339","article-title":"No subclass left behind: Finegrained robustness in coarse-grained classification problems","volume":"33","author":"Sohoni","year":"2020","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1033"},{"key":"ref47","first-page":"9229","article-title":"Test-time training with selfsupervision for generalization under distribution shifts","volume-title":"International conference on machine learning","author":"Sun"},{"key":"ref48","author":"Wah","year":"2011","journal-title":"The caltech-ucsd birds-200\u20132011 dataset"},{"key":"ref49","article-title":"Tent: Fully test-time adaptation by entropy minimization","volume-title":"International Conference on Learning Representations","author":"Wang"},{"key":"ref50","article-title":"Tent: Fully test-time adaptation by entropy minimization","author":"Wang","journal-title":"arXiv preprint"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr.2017.369"},{"key":"ref52","article-title":"Benchmarking deletion metrics with the principled explanations","volume-title":"Forty-first International Conference on Machine Learning","author":"Wang"},{"key":"ref53","first-page":"3745","article-title":"On the effect of key factors in spurious correlation: A theoretical perspective","volume-title":"International Conference on Artificial Intelligence and Statistics","author":"Wang"},{"key":"ref54","first-page":"37765","article-title":"Discover and cure: Concept-aware mitigation of spurious correlation","volume-title":"International Conference on Machine Learning","author":"Wu"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72952-2_13"},{"key":"ref56","first-page":"39365","article-title":"Mitigating spurious correlations in multimodal models during fine-tuning","volume-title":"International Conference on Machine Learning","author":"Yang"},{"key":"ref57","article-title":"Change is hard: A closer look at subpopulation shift","author":"Yang","year":"2023","journal-title":"arXiv preprint"},{"key":"ref58","first-page":"25407","article-title":"Improving out-of-distribution robustness via selective augmentation","volume-title":"International Conference on Machine Learning","author":"Yao"},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.02245"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.52202\/068431-2799"},{"key":"ref61","first-page":"26484","article-title":"Correct-n-contrast: a contrastive approach for improving robustness to spurious correlations","volume-title":"International Conference on Machine Learning","author":"Zhang"},{"key":"ref62","article-title":"Biomedclip: a multimodal biomedical foundation model pretrained from fifteen million scientific image-text pairs","author":"Zhang","year":"2023","journal-title":"arXiv preprint"},{"key":"ref63","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2017.2723009"}],"event":{"name":"2025 IEEE\/CVF International Conference on Computer Vision (ICCV)","location":"Honolulu, HI, USA","start":{"date-parts":[[2025,10,19]]},"end":{"date-parts":[[2025,10,25]]}},"container-title":["2025 IEEE\/CVF International Conference on Computer Vision (ICCV)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11443115\/11443287\/11445182.pdf?arnumber=11445182","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,30]],"date-time":"2026-04-30T06:14:19Z","timestamp":1777529659000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11445182\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,19]]},"references-count":63,"URL":"https:\/\/doi.org\/10.1109\/iccv51701.2025.00280","relation":{},"subject":[],"published":{"date-parts":[[2025,10,19]]}}}