{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,24]],"date-time":"2026-09-24T18:12:46Z","timestamp":1790273566644,"version":"4.1.0"},"reference-count":44,"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":"NSF","doi-asserted-by":"publisher","award":["2026498"],"award-info":[{"award-number":["2026498"]}],"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\/iccv51701.2025.02119","type":"proceedings-article","created":{"date-parts":[[2026,4,29]],"date-time":"2026-04-29T19:45:49Z","timestamp":1777491949000},"page":"22826-22835","source":"Crossref","is-referenced-by-count":4,"title":["Feather the Throttle: Revisiting Visual Token Pruning for Vision-Language Model Acceleration"],"prefix":"10.1109","author":[{"given":"Mark","family":"Endo","sequence":"first","affiliation":[{"name":"Stanford University,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiaohan","family":"Wang","sequence":"additional","affiliation":[{"name":"Stanford University,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Serena","family":"Yeung-Levy","sequence":"additional","affiliation":[{"name":"Stanford University,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v33i01.33018076"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1723"},{"key":"ref3","article-title":"Hired: Attention-guided token dropping for efficient inference of high-resolution vision-language models in resourceconstrained environments","author":"Arif","year":"2024","journal-title":"arXiv preprint"},{"key":"ref4","article-title":"Qwen-vl: A frontier large vision-language model with versatile abilities","author":"Bai","year":"2023","journal-title":"arXiv preprint"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1145\/1866029.1866080"},{"key":"ref6","article-title":"Token merging: Your vit but faster","volume-title":"The Eleventh International Conference on Learning Representations","author":"Bolya","year":"2023"},{"key":"ref7","article-title":"Matryoshka multimodal models","author":"Cai","year":"2024","journal-title":"arXiv preprint"},{"key":"ref8","article-title":"Auroracap: Efficient, performant video detailed captioning and a new benchmark","author":"Chai","year":"2024","journal-title":"arXiv preprint"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.52202\/079017-2353"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-73004-7_2"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v39i2.32229"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1189"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.670"},{"key":"ref14","article-title":"Zipvl: Efficient large vision-language models with dynamic token sparsification and kv cache compression","author":"He","year":"2024","journal-title":"arXiv preprint"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.findings-emnlp.338"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00686"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.01300"},{"key":"ref18","first-page":"23123","article-title":"Prismatic VLMs: Investigating the design space of visuallyconditioned language models","volume-title":"Proceedings of the 41st International Conference on Machine Learning","author":"Karamcheti","year":"2024"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.3115\/v1\/D14-1086"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-46493-0_15"},{"key":"ref21","first-page":"17044","article-title":"Naturalbench: Evaluating vision-language models on natural adversarial samples","volume":"37","author":"Li","year":"2025","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref22","first-page":"12888","article-title":"Blip: Bootstrapping language-image pre-training for unified vision-language understanding and generation","volume-title":"International conference on machine learning","author":"Li","year":"2022"},{"key":"ref23","first-page":"19730","article-title":"Blip-2: Bootstrapping language-image pre-training with frozen image encoders and large language models","volume-title":"International conference on machine learning","author":"Li","year":"2023"},{"key":"ref24","article-title":"Inference optimal vlms need only one visual token but larger models","author":"Li","year":"2024","journal-title":"arXiv preprint"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.emnlp-main.20"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1162\/tacl_a_00566"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.02484"},{"key":"ref28","author":"Liu","year":"2024","journal-title":"Llava-next: Improved reasoning, ocr, and world knowledge"},{"key":"ref29","article-title":"Visual instruction tuning","volume":"36","author":"Liu","year":"2024","journal-title":"Advances in neural information processing systems"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1109\/WACV57701.2024.00264"},{"key":"ref31","article-title":"Llava-prumerge: Adaptive token reduction for efficient large multimodal models","author":"Shang","year":"2024","journal-title":"arXiv preprint"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00851"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1016\/j.neucom.2023.127063"},{"key":"ref34","article-title":"Cambrian1: A fully open, vision-centric exploration of multimodal llms","author":"Tong","year":"2024","journal-title":"arXiv preprint"},{"key":"ref35","article-title":"Llama 2: Open foundation and fine-tuned chat models","author":"Touvron","year":"2023","journal-title":"arXiv preprint"},{"key":"ref36","article-title":"Look-m: Lookonce optimization in kv cache for efficient multimodal longcontext inference","author":"Wan","year":"2024","journal-title":"arXiv preprint"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.naacl-main.419"},{"key":"ref38","article-title":"Qwen2-vl: Enhancing vision-language model\u2019s perception of the world at any resolution","author":"Wang","year":"2024","journal-title":"arXiv preprint"},{"key":"ref39","article-title":"Pyramiddrop: Accelerating your large vision-language models via pyramid visual redundancy reduction","author":"Xing","year":"2024","journal-title":"arXiv preprint"},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52734.2025.01843"},{"key":"ref41","article-title":"Balancing performance and efficiency: A multimodal large language model pruning method based image text interaction","author":"Yu","year":"2024","journal-title":"arXiv preprint"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-46475-6_5"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.01100"},{"key":"ref44","first-page":"2412","article-title":"[cls] attention is all you need for training-free visual token pruning: Make vlm inference faster","author":"Zhang","year":"2024","journal-title":"arXiv eprints"}],"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\/11445089.pdf?arnumber=11445089","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,30]],"date-time":"2026-04-30T06:20:16Z","timestamp":1777530016000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11445089\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,19]]},"references-count":44,"URL":"https:\/\/doi.org\/10.1109\/iccv51701.2025.02119","relation":{},"subject":[],"published":{"date-parts":[[2025,10,19]]}}}