{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,6]],"date-time":"2026-05-06T07:10:43Z","timestamp":1778051443696,"version":"3.51.4"},"reference-count":41,"publisher":"IEEE","license":[{"start":{"date-parts":[[2026,3,6]],"date-time":"2026-03-06T00:00:00Z","timestamp":1772755200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,3,6]],"date-time":"2026-03-06T00:00:00Z","timestamp":1772755200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,3,6]]},"DOI":"10.1109\/wacv61042.2026.00585","type":"proceedings-article","created":{"date-parts":[[2026,5,5]],"date-time":"2026-05-05T19:59:32Z","timestamp":1778011172000},"page":"6049-6058","source":"Crossref","is-referenced-by-count":0,"title":["Understanding the Visual Projection Space of Multimodal LLMs"],"prefix":"10.1109","author":[{"given":"Sungheon","family":"Jeong","sequence":"first","affiliation":[{"name":"UC Irvine"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yoojeong","family":"Song","sequence":"additional","affiliation":[{"name":"Soonchunhyang University"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Hyungjoon","family":"Kim","sequence":"additional","affiliation":[{"name":"Changwon National University"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1723"},{"key":"ref2","article-title":"Language models are few-shot learners","author":"Brown","year":"2020","journal-title":"NeurIPS"},{"key":"ref3","article-title":"Sparks of artificial general intelligence: Early experiments with gpt-4","author":"Bubeck","year":"2023"},{"key":"ref4","article-title":"Microsoft coco captions: Data collection and evaluation server","author":"Chen","year":"2015","journal-title":"CVPR"},{"key":"ref5","article-title":"The subspace voyager: Rethinking expressiveness of generative language models","volume-title":"International Conference on Learning Representations","author":"Choromanski"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.52202\/075280-2142"},{"key":"ref7","article-title":"An image is worth 16x16 words: Transformers for image recognition at scale","author":"Dosovitskiy","year":"2021","journal-title":"ICLR"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1137\/S0895479895290954"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D19-1006"},{"key":"ref10","article-title":"Representation degeneration problem in training deep neural networks","author":"Gao","year":"2019","journal-title":"ICLR"},{"key":"ref11","article-title":"Dissecting massively multilingual models with language-neutral representations","author":"Geva","year":"2023","journal-title":"ACL Findings"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.670"},{"key":"ref13","first-page":"21","article-title":"Extended grassmann kernels for subspace-based learning","author":"Hamm","year":"2008","journal-title":"Advances in neural information processing systems"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1145\/1390156.1390204"},{"key":"ref15","article-title":"Factool: Factuality benchmarking of vision-language generative models","author":"Kim","year":"2023","journal-title":"ICCV"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.emnlp-main.733"},{"key":"ref17","article-title":"Otter: A multi-modal model with in-context instruction tuning","author":"Li","year":"2023"},{"key":"ref18","article-title":"Blip-2: Bootstrapping language-image pre-training with frozen image encoders and large language models","author":"Li","year":"2023"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1405.0312"},{"key":"ref20","article-title":"Visual instruction tuning","volume-title":"Conference on Computer Vision and Pattern Recognition (CVPR)","author":"Liu"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1516"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00986"},{"key":"ref23","article-title":"Univar: Unifying vision-language representation with univariate distribution alignment","author":"Luo","year":"2024"},{"key":"ref24","year":"2023","journal-title":"Gpt-4 technical report"},{"key":"ref25","article-title":"The intriguing role of linear layers in deep learning","author":"Papyan","year":"2020"},{"key":"ref26","article-title":"Kosmos-2: Grounding language models to see and talk","author":"Peng","year":"2023"},{"key":"ref27","article-title":"Kosmos2: Grounding large language models to see and talk","author":"Peng","year":"2023"},{"key":"ref28","article-title":"Learning transferable visual models from natural language supervision","volume-title":"International Conference on Machine Learning","author":"Radford"},{"key":"ref29","article-title":"Galactica: A large language model for science","author":"Taylor","year":"2022"},{"key":"ref30","article-title":"Llama: Open and efficient foundation language models","author":"Touvron","year":"2023"},{"key":"ref31","article-title":"Multimodal few-shot learning with frozen language models","author":"Tsimpoukelli","year":"2021","journal-title":"NeurIPS"},{"key":"ref32","article-title":"Causal mediation analysis for interpreting neural nlp: The case of gender bias","volume-title":"Conference on Empirical Methods in Natural Language Processing","author":"Vig"},{"key":"ref33","article-title":"Alignscore: [insert part of the title here]","volume-title":"Proceedings of the [Conference Name] (e.g., NeurIPS \/ ICML \/ CVPR \/ ACL \u2026)","author":"Wang"},{"key":"ref34","author":"Wang","year":"2023","journal-title":"What does a vision-language model do when it grounds language in vision?"},{"key":"ref35","author":"Wang","year":"2023","journal-title":"What makes good in-context examples for gpt models?"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D18-1166"},{"key":"ref37","article-title":"Sparse projection for vision-language models","author":"Yang","year":"2023"},{"key":"ref38","article-title":"Alignscore: Evaluating input-output alignment in multimodal language models","author":"Ye","year":"2023","journal-title":"CVPR"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1162\/tacl_a_00166"},{"key":"ref40","article-title":"On the conditional geometry of language models","author":"Zhou","year":"2023","journal-title":"ICLR"},{"key":"ref41","article-title":"Minigpt-4: Enhancing vision-language understanding with advanced large language models","author":"Zhu","year":"2023"}],"event":{"name":"2026 IEEE\/CVF Winter Conference on Applications of Computer Vision (WACV)","location":"Tucson, AZ, USA","start":{"date-parts":[[2026,3,6]]},"end":{"date-parts":[[2026,3,10]]}},"container-title":["2026 IEEE\/CVF Winter Conference on Applications of Computer Vision (WACV)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11491838\/11491925\/11491988.pdf?arnumber=11491988","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,6]],"date-time":"2026-05-06T06:11:48Z","timestamp":1778047908000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11491988\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3,6]]},"references-count":41,"URL":"https:\/\/doi.org\/10.1109\/wacv61042.2026.00585","relation":{},"subject":[],"published":{"date-parts":[[2026,3,6]]}}}