{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,16]],"date-time":"2026-06-16T14:45:29Z","timestamp":1781621129654,"version":"3.54.5"},"publisher-location":"New York, NY, USA","reference-count":40,"publisher":"ACM","content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2025,10,27]]},"DOI":"10.1145\/3746027.3755479","type":"proceedings-article","created":{"date-parts":[[2025,10,25]],"date-time":"2025-10-25T05:50:47Z","timestamp":1761371447000},"page":"10181-10189","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":3,"title":["Compute Only 16 Tokens in One Timestep: Accelerating Diffusion Transformers with Cluster-Driven Feature Caching"],"prefix":"10.1145","author":[{"ORCID":"https:\/\/orcid.org\/0009-0003-0228-4000","authenticated-orcid":false,"given":"Zhixin","family":"Zheng","sequence":"first","affiliation":[{"name":"Shanghai Jiao Tong University, Shanghai, China and University of Electronic Science and Technology of China, Chengdu, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0003-5492-8015","authenticated-orcid":false,"given":"Xinyu","family":"Wang","sequence":"additional","affiliation":[{"name":"Shanghai Jiao Tong University, Shanghai, China and Shandong University, Qingdao, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-2663-2204","authenticated-orcid":false,"given":"Chang","family":"Zou","sequence":"additional","affiliation":[{"name":"University of Electronic Science and Technology of China, Chengdu, China and Shanghai Jiao Tong University, Shanghai, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8156-7081","authenticated-orcid":false,"given":"Shaobo","family":"Wang","sequence":"additional","affiliation":[{"name":"Shanghai Jiao Tong University, Shanghai, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3341-183X","authenticated-orcid":false,"given":"Linfeng","family":"Zhang","sequence":"additional","affiliation":[{"name":"Shanghai Jiao Tong University, Shanghai, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"320","published-online":{"date-parts":[[2025,10,27]]},"reference":[{"key":"e_1_3_2_1_1_1","unstructured":"Andreas Blattmann Tim Dockhorn Sumith Kulal Daniel Mendelevitch Maciej Kilian Dominik Lorenz Yam Levi Zion English Vikram Voleti Adam Letts et al. 2023. Stable video diffusion: Scaling latent video diffusion models to large datasets. arXiv preprint arXiv:2311.15127 (2023)."},{"key":"e_1_3_2_1_2_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW59228.2023.00484"},{"key":"e_1_3_2_1_3_1","volume-title":"\u0394-DiT: A Training-Free Acceleration Method Tailored for Diffusion Transformers. arXiv preprint arXiv:2406.01125","author":"Chen Pengtao","year":"2024","unstructured":"Pengtao Chen, Mingzhu Shen, Peng Ye, Jianjian Cao, Chongjun Tu, Christos-Savvas Bouganis, Yiren Zhao, and Tao Chen. 2024. \u0394-DiT: A Training-Free Acceleration Method Tailored for Diffusion Transformers. arXiv preprint arXiv:2406.01125 (2024)."},{"key":"e_1_3_2_1_4_1","unstructured":"Xinle Cheng Zhuoming Chen and Zhihao Jia. 2025. CAT Pruning: Cluster-Aware Token Pruning For Text-to-Image Diffusion Models. arXiv:2502.00433 [cs.CV] https:\/\/arxiv.org\/abs\/2502.00433"},{"key":"e_1_3_2_1_5_1","volume-title":"Flashattention: Fast and memory-efficient exact attention with io-awareness. Advances in neural information processing systems","author":"Dao Tri","year":"2022","unstructured":"Tri Dao, Dan Fu, Stefano Ermon, Atri Rudra, and Christopher R\u00e9. 2022. Flashattention: Fast and memory-efficient exact attention with io-awareness. Advances in neural information processing systems, Vol. 35 (2022), 16344-16359."},{"key":"e_1_3_2_1_6_1","volume-title":"Diffusion models beat gans on image synthesis. Advances in neural information processing systems","author":"Dhariwal Prafulla","year":"2021","unstructured":"Prafulla Dhariwal and Alexander Nichol. 2021. Diffusion models beat gans on image synthesis. Advances in neural information processing systems, Vol. 34 (2021), 8780-8794."},{"key":"e_1_3_2_1_7_1","unstructured":"Gongfan Fang Xinyin Ma and Xinchao Wang. 2023. Structural Pruning for Diffusion Models. arXiv:2305.10924 [cs.LG] https:\/\/arxiv.org\/abs\/2305.10924"},{"key":"e_1_3_2_1_8_1","volume-title":"Ronan Le Bras, and Yejin Choi","author":"Hessel Jack","year":"2021","unstructured":"Jack Hessel, Ari Holtzman, Maxwell Forbes, Ronan Le Bras, and Yejin Choi. 2021. Clipscore: A reference-free evaluation metric for image captioning. arXiv preprint arXiv:2104.08718 (2021)."},{"key":"e_1_3_2_1_9_1","volume-title":"Advances in Neural Information Processing Systems 30: Annual Conference on Neural Information Processing Systems 2017","author":"Heusel Martin","year":"2017","unstructured":"Martin Heusel, Hubert Ramsauer, Thomas Unterthiner, Bernhard Nessler, and Sepp Hochreiter. 2017. GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium. In Advances in Neural Information Processing Systems 30: Annual Conference on Neural Information Processing Systems 2017, December 4-9, 2017, Long Beach, CA, USA, Isabelle Guyon, Ulrike von Luxburg, Samy Bengio, Hanna M. Wallach, Rob Fergus, S. V. N. Vishwanathan, and Roman Garnett (Eds.). 6626-6637. https:\/\/proceedings.neurips.cc\/paper\/2017\/hash\/8a1d694707eb0fefe65871369074926d-Abstract.html"},{"key":"e_1_3_2_1_10_1","volume-title":"Denoising diffusion probabilistic models. Advances in neural information processing systems","author":"Ho Jonathan","year":"2020","unstructured":"Jonathan Ho, Ajay Jain, and Pieter Abbeel. 2020. Denoising diffusion probabilistic models. Advances in neural information processing systems, Vol. 33 (2020), 6840-6851."},{"key":"e_1_3_2_1_11_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.02060"},{"key":"e_1_3_2_1_12_1","doi-asserted-by":"publisher","DOI":"10.1109\/WACV57701.2024.00141"},{"key":"e_1_3_2_1_13_1","doi-asserted-by":"publisher","DOI":"10.1109\/HPCA61900.2025.00035"},{"key":"e_1_3_2_1_14_1","unstructured":"Black Forest Labs. 2024. FLUX. https:\/\/github.com\/black-forest-labs\/flux."},{"key":"e_1_3_2_1_15_1","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.01608"},{"key":"e_1_3_2_1_16_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52734.2025.00689"},{"key":"e_1_3_2_1_17_1","volume-title":"From reusing to forecasting: Accelerating diffusion models with taylorseers. arXiv preprint arXiv:2503.06923","author":"Liu Jiacheng","year":"2025","unstructured":"Jiacheng Liu, Chang Zou, Yuanhuiyi Lyu, Junjie Chen, and Linfeng Zhang. 2025b. From reusing to forecasting: Accelerating diffusion models with taylorseers. arXiv preprint arXiv:2503.06923 (2025)."},{"key":"e_1_3_2_1_18_1","volume-title":"From reusing to forecasting: Accelerating diffusion models with taylorseers. arXiv preprint arXiv:2503.06923","author":"Liu Jiacheng","year":"2025","unstructured":"Jiacheng Liu, Chang Zou, Yuanhuiyi Lyu, Junjie Chen, and Linfeng Zhang. 2025c. From reusing to forecasting: Accelerating diffusion models with taylorseers. arXiv preprint arXiv:2503.06923 (2025)."},{"key":"e_1_3_2_1_19_1","volume-title":"Flow straight and fast: Learning to generate and transfer data with rectified flow. arXiv preprint arXiv:2209.03003","author":"Liu Xingchao","year":"2022","unstructured":"Xingchao Liu, Chengyue Gong, and Qiang Liu. 2022. Flow straight and fast: Learning to generate and transfer data with rectified flow. arXiv preprint arXiv:2209.03003 (2022)."},{"key":"e_1_3_2_1_20_1","volume-title":"Dpm-solver: A fast ode solver for diffusion probabilistic model sampling in around 10 steps. Advances in neural information processing systems","author":"Lu Cheng","year":"2022","unstructured":"Cheng Lu, Yuhao Zhou, Fan Bao, Jianfei Chen, Chongxuan Li, and Jun Zhu. 2022. Dpm-solver: A fast ode solver for diffusion probabilistic model sampling in around 10 steps. Advances in neural information processing systems, Vol. 35 (2022), 5775-5787."},{"key":"e_1_3_2_1_21_1","volume-title":"Dpm-solver: Fast solver for guided sampling of diffusion probabilistic models. Machine Intelligence Research","author":"Lu Cheng","year":"2025","unstructured":"Cheng Lu, Yuhao Zhou, Fan Bao, Jianfei Chen, Chongxuan Li, and Jun Zhu. 2025. Dpm-solver: Fast solver for guided sampling of diffusion probabilistic models. Machine Intelligence Research (2025), 1-22."},{"key":"e_1_3_2_1_22_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.01492"},{"key":"e_1_3_2_1_23_1","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.00387"},{"key":"e_1_3_2_1_24_1","doi-asserted-by":"publisher","DOI":"10.1007\/S11263-015-0816-Y"},{"key":"e_1_3_2_1_25_1","volume-title":"Atiyeh Gh Moghadam, and Ahmad Nickabadi","author":"Saghatchian Omid","year":"2025","unstructured":"Omid Saghatchian, Atiyeh Gh Moghadam, and Ahmad Nickabadi. 2025. Cached Adaptive Token Merging: Dynamic Token Reduction and Redundant Computation Elimination in Diffusion Model. arXiv preprint arXiv:2501.00946 (2025)."},{"key":"e_1_3_2_1_26_1","volume-title":"Burcu Karagol Ayan, Tim Salimans, et al.","author":"Saharia Chitwan","year":"2022","unstructured":"Chitwan Saharia, William Chan, Saurabh Saxena, Lala Li, Jay Whang, Emily L Denton, Kamyar Ghasemipour, Raphael Gontijo Lopes, Burcu Karagol Ayan, Tim Salimans, et al., 2022. Photorealistic text-to-image diffusion models with deep language understanding. Advances in neural information processing systems, Vol. 35 (2022), 36479-36494."},{"key":"e_1_3_2_1_27_1","volume-title":"Fora: Fast-forward caching in diffusion transformer acceleration. arXiv preprint arXiv:2407.01425","author":"Selvaraju Pratheba","year":"2024","unstructured":"Pratheba Selvaraju, Tianyu Ding, Tianyi Chen, Ilya Zharkov, and Luming Liang. 2024. Fora: Fast-forward caching in diffusion transformer acceleration. arXiv preprint arXiv:2407.01425 (2024)."},{"key":"e_1_3_2_1_28_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.00196"},{"key":"e_1_3_2_1_29_1","volume-title":"Denoising diffusion implicit models. arXiv preprint arXiv:2010.02502","author":"Song Jiaming","year":"2020","unstructured":"Jiaming Song, Chenlin Meng, and Stefano Ermon. 2020. Denoising diffusion implicit models. arXiv preprint arXiv:2010.02502 (2020)."},{"key":"e_1_3_2_1_30_1","unstructured":"Yang Song Prafulla Dhariwal Mark Chen and Ilya Sutskever. 2023. Consistency models. (2023)."},{"key":"e_1_3_2_1_31_1","volume-title":"Hunyuan-large: An open-source moe model with 52 billion activated parameters by tencent. arXiv preprint arXiv:2411.02265","author":"Sun Xingwu","year":"2024","unstructured":"Xingwu Sun, Yanfeng Chen, Yiqing Huang, Ruobing Xie, Jiaqi Zhu, Kai Zhang, Shuaipeng Li, Zhen Yang, Jonny Han, Xiaobo Shu, et al., 2024. Hunyuan-large: An open-source moe model with 52 billion activated parameters by tencent. arXiv preprint arXiv:2411.02265 (2024)."},{"key":"e_1_3_2_1_32_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.00594"},{"key":"e_1_3_2_1_33_1","first-page":"15903","article-title":"Imagereward: Learning and evaluating human preferences for text-to-image generation","volume":"36","author":"Xu Jiazheng","year":"2023","unstructured":"Jiazheng Xu, Xiao Liu, Yuchen Wu, Yuxuan Tong, Qinkai Li, Ming Ding, Jie Tang, and Yuxiao Dong. 2023. Imagereward: Learning and evaluating human preferences for text-to-image generation. Advances in Neural Information Processing Systems, Vol. 36 (2023), 15903-15935.","journal-title":"Advances in Neural Information Processing Systems"},{"key":"e_1_3_2_1_34_1","first-page":"1196","article-title":"Ditfastattn: Attention compression for diffusion transformer models","volume":"37","author":"Yuan Zhihang","year":"2024","unstructured":"Zhihang Yuan, Hanling Zhang, Lu Pu, Xuefei Ning, Linfeng Zhang, Tianchen Zhao, Shengen Yan, Guohao Dai, and Yu Wang. 2024. Ditfastattn: Attention compression for diffusion transformer models. Advances in Neural Information Processing Systems, Vol. 37 (2024), 1196-1219.","journal-title":"Advances in Neural Information Processing Systems"},{"key":"e_1_3_2_1_35_1","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v39i9.33071"},{"key":"e_1_3_2_1_36_1","volume-title":"Token pruning for caching better: 9 times acceleration on stable diffusion for free. arXiv preprint arXiv:2501.00375","author":"Zhang Evelyn","year":"2024","unstructured":"Evelyn Zhang, Bang Xiao, Jiayi Tang, Qianli Ma, Chang Zou, Xuefei Ning, Xuming Hu, and Linfeng Zhang. 2024. Token pruning for caching better: 9 times acceleration on stable diffusion for free. arXiv preprint arXiv:2501.00375 (2024)."},{"key":"e_1_3_2_1_37_1","volume-title":"Real-time video generation with pyramid attention broadcast. arXiv preprint arXiv:2408.12588","author":"Zhao Xuanlei","year":"2024","unstructured":"Xuanlei Zhao, Xiaolong Jin, Kai Wang, and Yang You. 2024. Real-time video generation with pyramid attention broadcast. arXiv preprint arXiv:2408.12588 (2024)."},{"key":"e_1_3_2_1_38_1","first-page":"92581","article-title":"Dip-go: A diffusion pruner via few-step gradient optimization","volume":"37","author":"Zhu Haowei","year":"2024","unstructured":"Haowei Zhu, Dehua Tang, Ji Liu, Mingjie Lu, Jintu Zheng, Jinzhang Peng, Dong Li, Yu Wang, Fan Jiang, Lu Tian, et al., 2024. Dip-go: A diffusion pruner via few-step gradient optimization. Advances in Neural Information Processing Systems, Vol. 37 (2024), 92581-92604.","journal-title":"Advances in Neural Information Processing Systems"},{"key":"e_1_3_2_1_39_1","volume-title":"Accelerating diffusion transformers with token-wise feature caching. arXiv preprint arXiv:2410.05317","author":"Zou Chang","year":"2024","unstructured":"Chang Zou, Xuyang Liu, Ting Liu, Siteng Huang, and Linfeng Zhang. 2024a. Accelerating diffusion transformers with token-wise feature caching. arXiv preprint arXiv:2410.05317 (2024)."},{"key":"e_1_3_2_1_40_1","volume-title":"Accelerating diffusion transformers with dual feature caching. arXiv preprint arXiv:2412.18911","author":"Zou Chang","year":"2024","unstructured":"Chang Zou, Evelyn Zhang, Runlin Guo, Haohang Xu, Conghui He, Xuming Hu, and Linfeng Zhang. 2024b. Accelerating diffusion transformers with dual feature caching. arXiv preprint arXiv:2412.18911 (2024)."}],"event":{"name":"MM '25: The 33rd ACM International Conference on Multimedia","location":"Dublin Ireland","acronym":"MM '25","sponsor":["SIGMM ACM Special Interest Group on Multimedia"]},"container-title":["Proceedings of the 33rd ACM International Conference on Multimedia"],"original-title":[],"link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3746027.3755479","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,12,9]],"date-time":"2025-12-09T19:25:17Z","timestamp":1765308317000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3746027.3755479"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,27]]},"references-count":40,"alternative-id":["10.1145\/3746027.3755479","10.1145\/3746027"],"URL":"https:\/\/doi.org\/10.1145\/3746027.3755479","relation":{},"subject":[],"published":{"date-parts":[[2025,10,27]]},"assertion":[{"value":"2025-10-27","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}