发表机构
Manifold AI; Tsinghua University（流形人工智能; 清华大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
DensityKV是一种无需训练的历史KV库管理策略，通过密度引导压缩KV缓存，提升长视频生成的时序一致性与稳定性，且存储不随推演长度增长。
AI 中文摘要
自回归视频扩散模型通过滑动窗口注意力实现流式生成，但每个生成块依赖于先前生成的内容，导致外观和运动误差随时间递归传播。历史键值（KV）内存保留了早期主体和场景状态，有助于维持长时序一致性。然而，保留每个生成状态会形成随推演不断增长的历史存档，而重复状态会反复添加冗余覆盖。为解决该问题，我们提出DensityKV，一种无需训练的历史KV库管理策略。DensityKV为每个注意力头维护单独的令牌级KV库，并使用Soft-Riesz密度测量直接参数化注意力路由的RoPE后键的局部冗余。通过约束状态进入库后邻域密度的增长，DensityKV在保留每个完成生成块的连贯状态的同时，限制了重复的历史积累。在三种自回归视频生成主干和多种生成长度上的实验表明，在历史KV容量上限相同的情况下，DensityKV提升了长时序一致性和生成稳定性，同时保持了独立于推演长度的持久历史存储。
英文摘要
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.
Comments17 pages, 9 figures, 2 tables. Code: https://github.com/ZhaoWQQ/DensityKV
