{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,13]],"date-time":"2026-05-13T19:12:43Z","timestamp":1778699563713,"version":"3.51.4"},"reference-count":114,"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\/100000005","name":"DoD","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100000005","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\/iccv51701.2025.00265","type":"proceedings-article","created":{"date-parts":[[2026,4,29]],"date-time":"2026-04-29T19:45:49Z","timestamp":1777491949000},"page":"2760-2772","source":"Crossref","is-referenced-by-count":1,"title":["Enhancing Few-Shot Vision-Language Classification With Large Multimodal Model Features"],"prefix":"10.1109","author":[{"given":"Chancharik","family":"Mitra","sequence":"first","affiliation":[{"name":"Carnegie Mellon University"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Brandon","family":"Huang","sequence":"additional","affiliation":[{"name":"University of California,Berkeley"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Tianning","family":"Chai","sequence":"additional","affiliation":[{"name":"University of California,Berkeley"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Zhiqiu","family":"Lin","sequence":"additional","affiliation":[{"name":"Carnegie Mellon University"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Assaf","family":"Arbelle","sequence":"additional","affiliation":[{"name":"IBM Research"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Rogerio","family":"Feris","sequence":"additional","affiliation":[{"name":"MIT-IBM Watson AI Lab"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Leonid","family":"Karlinsky","sequence":"additional","affiliation":[{"name":"MIT-IBM Watson AI Lab"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Trevor","family":"Darrell","sequence":"additional","affiliation":[{"name":"University of California,Berkeley"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Deva","family":"Ramanan","sequence":"additional","affiliation":[{"name":"Carnegie Mellon University"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Roei","family":"Herzig","sequence":"additional","affiliation":[{"name":"University of California,Berkeley"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","article-title":"A broad-coverage challenge corpus for sentence understanding through inference","author":"Bowman","year":"2017","journal-title":"ArXiv"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2015.279"},{"key":"ref3","article-title":"Qwen-vl: A frontier large vision-language model with versatile abilities","author":"Bai","year":"2023","journal-title":"ArXiv"},{"key":"ref4","article-title":"Autoencoders, unsupervised learning, and deep architectures","volume-title":"ICML 2011 Unsupervised and Transfer Learning Workshop","author":"Baldi","year":"2011"},{"key":"ref5","volume-title":"Introducing our multimodal models","author":"Bavishi","year":"2023"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.trustnlp-1.5"},{"key":"ref7","article-title":"Language models are few-shot learners","author":"Brown","year":"2020","journal-title":"ArXiv"},{"key":"ref8","article-title":"Finetuned \u2019small\u2019 11 ms (still) significantly outperform zeroshot generative ai models in text classification","author":"Bucher","year":"2024","journal-title":"ArXiv"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.acl-long.178"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2009.5206848"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1145\/3543873.3587605"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.01101"},{"key":"ref13","article-title":"Bert: Pre-training of deep bidirectional transformers for language understanding","author":"Devlin","year":"2019","journal-title":"ArXiv"},{"key":"ref14","article-title":"Towards multimodal in-context learning for vision & language models","author":"Doveh","year":"2024","journal-title":"ArXiv"},{"key":"ref15","article-title":"Training vision transformers for image retrieval","author":"El-Nouby","year":"2021","journal-title":"ArXiv"},{"key":"ref16","article-title":"Kto: Model alignment as prospect theoretic optimization","author":"Ethayarajh","year":"2024","journal-title":"ArXiv"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.1112937108"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-73337-6_9"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.emnlp-main.552"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00380"},{"key":"ref21","article-title":"Merlin: Multimodal embedding refinement via 11 m -based iterative navigation for text-video retrieval-rerank pipeline","author":"Han","year":"2024","journal-title":"arxiv preprint arxiv"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/JSTARS.2019.2918242"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.findings-emnlp.624"},{"key":"ref24","article-title":"Incontext learning creates task vectors","author":"Hendel","year":"2023","journal-title":"arxiv preprint arxiv"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.emnlp-main.870"},{"key":"ref26","article-title":"Visualizing data using t-sne","author":"G","year":"2008","journal-title":"Journal of machine learning research"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72775-7_15"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1355"},{"key":"ref29","article-title":"Lora: Low-rank adaptation of large language models","author":"Hu","year":"2021","journal-title":"arxiv preprint arxiv"},{"key":"ref30","article-title":"Multimodal task vectors enable many-shot multimodal in-context learning","author":"Huang","year":"2024","journal-title":"arxiv preprint arxiv"},{"key":"ref31","article-title":"Llm2clip: Powerful language model unlock richer visual representation","author":"Huang","year":"2024","journal-title":"arxiv preprint arxiv"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00686"},{"key":"ref33","article-title":"Learning object detection from captions via textual scene attributes","author":"Jerbi","year":"2020","journal-title":"ArXiv"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.emnlp-main.603"},{"key":"ref35","article-title":"Scaling sentence embeddings with large language models","author":"Jiang","year":"2023","journal-title":"ArXiv"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.findings-emnlp.181"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1038\/77664"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.3115\/v1\/d14-1086"},{"key":"ref39","article-title":"and Max Welling. Auto-encoding variational bayes","author":"Kingma","year":"2013","journal-title":"CoRR"},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01123"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.acl-long.546"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.emnlp-main.243"},{"key":"ref43","article-title":"Seed-bench: Benchmarking multimodal 11 ms with generative comprehension","author":"Li","year":"2023","journal-title":"ArXiv"},{"key":"ref44","article-title":"Naturalbench: Evaluating vision-language models on natural adversarial samples","author":"Li","year":"2024","journal-title":"arxiv preprint arxiv"},{"key":"ref45","article-title":"Llava-onevision: Easy visual task transfer","author":"Li","year":"2024","journal-title":"ArXiv"},{"key":"ref46","article-title":"Blip: Bootstrapping language-image pre-training for unified vision-language understanding and generation","author":"Li","year":"2022","journal-title":"arxiv preprint arxiv"},{"key":"ref47","article-title":"BLIP-2: bootstrapping language-image pre-training with frozen image encoders and large language models","author":"Li","year":"2023","journal-title":"ICML"},{"key":"ref48","article-title":"Conanembedding: General text embedding with more and better negative samples","author":"Li","year":"2024","journal-title":"ArXiv"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.acllong.353"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52729.2023.02240"},{"key":"ref51","article-title":"Your mixture-of-experts llm is secretly an embedding model for free","author":"Li","year":"2024","journal-title":"arxiv preprint arxiv"},{"key":"ref52","article-title":"Your mixture-of-experts llm is secretly an embedding model for free","author":"Li","year":"2024","journal-title":"arxiv preprint arxiv"},{"key":"ref53","article-title":"Unleashing the power of 11 ms as multi-modal encoders for text and graphstructured data","author":"Lin","year":"2024","journal-title":"arxiv preprint arxiv"},{"key":"ref54","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1405.0312"},{"key":"ref55","article-title":"Revisiting the role of language priors in vision-language models","author":"Lin","year":"2023","journal-title":"arxiv preprint arxiv"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.01852"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72673-6_20"},{"key":"ref58","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72673-6_20"},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52733.2024.02484"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1516"},{"key":"ref61","article-title":"Meaning representations from trajectories in autoregressive models","author":"Yu Liu","year":"2023","journal-title":"ArXiv"},{"key":"ref62","article-title":"Decision-making with auto-encoding variational bayes","author":"Lopez","year":"2020","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref63","first-page":"5081","article-title":"Decision-making with auto-encoding variational bayes","volume":"33","author":"Lopez","year":"2020","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref64","article-title":"Ovis: Structural embedding alignment for multimodal large language model","author":"Lu","year":"2024","journal-title":"ArXiv"},{"key":"ref65","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr.2016.9"},{"key":"ref66","article-title":"Corrado, and Jeff Dean. Distributed representations of words and phrases and their compositionality","author":"Mikolov","year":"2013","journal-title":"Advances in neural information processing systems"},{"key":"ref67","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.01367"},{"key":"ref68","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.eacl-main.148"},{"key":"ref69","article-title":"Sentence-t5: Scalable sentence encoders from pre-trained text-to-text models","author":"Ni","year":"2021","journal-title":"ArXiv"},{"key":"ref70","article-title":"Large dual encoders are generalizable retrievers","author":"Ni","year":"2021","journal-title":"ArXiv"},{"key":"ref71","doi-asserted-by":"publisher","DOI":"10.1109\/ICVGIP.2008.47"},{"key":"ref72","article-title":"Incontext learning and induction heads","author":"Olsson","year":"2022","journal-title":"arxiv preprint arxiv"},{"key":"ref73","article-title":"Gpt-4 technical report","year":"2023","journal-title":"ArXiv"},{"key":"ref74","article-title":"Training language models to follow instructions with human feedback","author":"Ouyang","year":"2022","journal-title":"ArXiv"},{"key":"ref75","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2012.6248092"},{"key":"ref76","article-title":"Pytorch: An imperative style, high-performance deep learning library","author":"Paszke","year":"2019","journal-title":"Advances in neural information processing systems, 32"},{"key":"ref77","doi-asserted-by":"publisher","DOI":"10.3115\/v1\/D14-1162"},{"key":"ref78","first-page":"8748","article-title":"Learning transferable visual models from natural language supervision","volume-title":"International Conference on Machine Learning (ICML)","author":"Radford","year":"2021"},{"key":"ref79","article-title":"Direct preference optimization: Your language model is secretly a reward model","author":"Rafailov","year":"2024","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref80","article-title":"Hierarchical text-conditional image generation with clip latents","author":"Ramesh","year":"2022","journal-title":"ArXiv"},{"key":"ref81","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D19-1410"},{"key":"ref82","doi-asserted-by":"publisher","DOI":"10.7551\/mitpress\/5236.001.0001"},{"key":"ref83","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2015.7298682"},{"key":"ref84","doi-asserted-by":"publisher","DOI":"10.1017\/CBO9781107298019"},{"key":"ref85","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00851"},{"key":"ref86","article-title":"The transient nature of emergent in-context learning in transformers","author":"Singh","year":"2023","journal-title":"Advances in neural information processing systems (NeurIPS)"},{"key":"ref87","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D13-1170"},{"key":"ref88","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.findings-emnlp.603"},{"key":"ref89","article-title":"Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context","year":"2024","journal-title":"ArXiv"},{"key":"ref90","article-title":"Gemini: a family of highly capable multimodal models","author":"Team","year":"2023","journal-title":"arxiv preprint arxiv"},{"key":"ref91","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.00517"},{"key":"ref92","article-title":"Wallace, and David Bau. Function vectors in large language models","author":"Todd","year":"2023","journal-title":"ArXiv"},{"key":"ref93","article-title":"Wallace, and David Bau. Function vectors in large language models","author":"Todd","year":"2023","journal-title":"arxiv preprint arxiv"},{"key":"ref94","doi-asserted-by":"publisher","DOI":"10.1162\/jocn.1991.3.1.71"},{"key":"ref95","article-title":"Neural discrete representation learning","author":"van den Oord","year":"2017","journal-title":"ArXiv"},{"key":"ref96","volume-title":"The caltech-ucsd birds-2002011 dataset","author":"Wah","year":"2011"},{"key":"ref97","article-title":"Improving text embeddings with large language models","author":"Wang","year":"2023","journal-title":"ArXiv"},{"key":"ref98","article-title":"Qwen2-vl: Enhancing vision-language model\u2019s perception of the world at any resolution","author":"Wang","year":"2024","journal-title":"ArXiv"},{"key":"ref99","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.emnlp-main.585"},{"key":"ref100","article-title":"Emu3: Next-token prediction is all you need","author":"Wang","year":"2024","journal-title":"arxiv preprint arxiv"},{"key":"ref101","article-title":"Internvideo2: Scaling video foundation models for multimodal video understanding","author":"Wang","year":"2024","journal-title":"ArXiv"},{"key":"ref102","article-title":"Large language models are zero-shot text classifiers","author":"Wang","year":"2023","journal-title":"ArXiv"},{"key":"ref103","article-title":"Finetuned language models are zero-shot learners","author":"Wei","year":"2021","journal-title":"ArXiv"},{"key":"ref104","article-title":"Chain of thought prompting elicits reasoning in large language models","author":"Wei","year":"2022","journal-title":"ArXiv"},{"key":"ref105","doi-asserted-by":"publisher","DOI":"10.1002\/ail2.109"},{"key":"ref106","article-title":"mplug-owl: Modularization empowers large language models with multimodality","author":"Ye","year":"2023","journal-title":"ArXiv"},{"key":"ref107","doi-asserted-by":"publisher","DOI":"10.1162\/tacl_a_00166"},{"key":"ref108","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.01100"},{"key":"ref109","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.01100"},{"key":"ref110","article-title":"Contrastive learning of medical visual representations from paired images and text","author":"Zhang","year":"2020","journal-title":"Machine Learning in Health Care"},{"key":"ref111","article-title":"Why are visually-grounded language models bad at image classification","author":"Zhang","year":"2024","journal-title":"ArXiv"},{"key":"ref112","article-title":"Mmicl: Empowering vision-language model with multi-modal in-context learning","author":"Zhao","year":"2023","journal-title":"ArXiv"},{"key":"ref113","article-title":"Minigpt-4: Enhancing vision-language understanding with advanced large language models","author":"Zhu","year":"2023","journal-title":"arxiv preprint arxiv"},{"key":"ref114","article-title":"Safety fine-tuning at (almost) no cost: A baseline for vision large language models","author":"Zong","year":"2024","journal-title":"arxiv preprint arxiv"}],"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\/11444922.pdf?arnumber=11444922","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T05:00:31Z","timestamp":1777611631000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11444922\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,19]]},"references-count":114,"URL":"https:\/\/doi.org\/10.1109\/iccv51701.2025.00265","relation":{},"subject":[],"published":{"date-parts":[[2025,10,19]]}}}