DensityKV: Density-Guided KV Cache Compression for Long Video Generation

[2608.27922] DensityKV: Density-Guided KV Cache Compression for Long Video Generation Skip to main content Search Submit Donate Log in Computer Science > Computer Vision and Pattern Recognition arXiv:2608.27922 (cs) [Submitted on 28 Aug 2026] Title: DensityKV: Density-Guided KV Cache Compression for Long Video Generation Authors: Wenqu Zhao , Xuemin Chi , Xin Zhang , Guoqing Ma , Baorun Li , Jianjie Fang , Peizhi Tang , Chen Gao , Wei Wu View PDF HTML (experimental) Abstract: Autoregressive video diffusion models enable streaming generation through sliding-window attention, but each generated block is conditioned on previously generated content, causing appearance and motion errors to propagate recursively over time. Historical key-value (KV) memory preserves earlier subject and scene states and helps maintain long-horizon consistency. However, retaining every generated state creates a historical archive that grows continuously with the rollout, while recurrent states repeatedly add redundant coverage. To address this problem, we propose DensityKV, a training-free historical KV bank management strategy. DensityKV maintains a separate token-level KV bank for each attention head and measures local redundancy among the post-RoPE keys that directly parameterize attention routing using Soft-Riesz density. By constraining neighborhood-density growth after states enter the bank, DensityKV limits repeated historical accumulation while preserving coherent states from each completed generation block. Experiments across three autoregressive video generation backbones and multiple generation lengths show that, at the same upper bound on historical KV capacity, DensityKV improves long-horizon consistency and generation stability while keeping persistent historical storage bounded independently of rollout length. Comments: 17 pages, 9 figures, 2 tables. Code: this https URL Subjects: Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2608.27922 [cs.CV] (or arXiv:2608.27922v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2608.27922 Focus to learn more arXiv-issued DOI via DataCite Submission history From: Wenqu Zhao [ view email ] [v1] Fri, 28 Aug 2026 04:57:57 UTC (19,289 KB) Full-text links: Access Paper: View PDF HTML (experimental) TeX Source view license Additional Features Audio Summary Current browse context: cs.CV < prev | next > new | recent | 2026-08 Change to browse by: cs References & Citations SciX Google Scholar Semantic Scholar export BibTeX citation Bookmark Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer ( What is the Explorer? ) Connected Papers Toggle Connected Papers ( What is Connected Papers? ) Litmaps Toggle Litmaps ( What is Litmaps? ) scite.ai Toggle scite Smart Citations ( What are Smart Citations? ) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv ( What is alphaXiv? ) Links to Code Toggle CatalyzeX Code Finder for Papers ( What is CatalyzeX? ) DagsHub Toggle DagsHub ( What is DagsHub? ) GotitPub Toggle Gotit.pub ( What is GotitPub? ) Huggingface Toggle Hugging Face ( What is Huggingface? ) ScienceCast Toggle ScienceCast ( What is ScienceCast? ) Demos Demos Replicate Toggle Replicate ( What is Replicate? ) Spaces Toggle Hugging Face Spaces ( What is Spaces? ) Spaces Toggle TXYZ.AI ( What is TXYZ.AI? ) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower ( What are Influence Flowers? ) Core recommender toggle CORE Recommender ( What is CORE? ) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs . Which authors of this paper are endorsers? | Disable MathJax ( What is MathJax? ) We gratefully acknowledge support from our major funders , member institutions , and all contributors. About · Help · Contact · Subscribe · Copyright · Privacy · Accessibility · Operational Status (opens in new tab)

Related documents

• DensityKV: Density-Guided KV Cache Compression for Long Video Generation#779825
• LayerRecall: A State-Conditioned Memory Router for Long-Horizon Consistency in Video Generation#792464
• LayerRecall: A State-Conditioned Memory Router for Long-Horizon Consistency in Video Generation#788464
• PREM: Prefix-Steered Recurrent Memory for Long-Video Understanding#1049340
• Teaching Video Generators to Remember: Eliciting Dynamic Memory for Out-of-Sight State Evolution#1131548