paper-with-me

홈 › Papers

DynamicRad: Content-Adaptive Sparse Attention for Long Video Diffusion

2026-04-22 · Yongji Long, Shijun Liang, Jintao Li, Yun Li arxiv

Leveraging the natural spatiotemporal energy decay in video diffusion offers a path to efficiency, yet relying solely on rigid static masks risks losing critical long-range information in complex dynamics. To address this issue, we propose \textbf{DynamicRad}, a unified sparse-attention paradigm that grounds adaptive selection within a radial locality prior. DynamicRad introduces a \textbf{dual-mode} strategy: \textit{static-ratio} for speed-optimized execution and \textit{dynamic-threshold} for quality-first filtering. To ensure robustness without online search overhead, we integrate an offline Bayesian Optimization (BO) pipeline coupled with a \textbf{semantic motion router}. This lightweight projection module maps prompt embeddings to optimal sparsity regimes with \textbf{minimal runtime overhead}. Unlike online profiling methods, our offline BO optimizes attention reconstruction error (MSE) on a physics-based proxy task, ensuring rapid convergence. Experiments on HunyuanVideo and Wan2.1-14B demonstrate that DynamicRad pushes the efficiency--quality Pareto frontier, achieving \textbf{1.7$\times$--2.5$\times$ inference speedups} with \textbf{over 80\% effective sparsity}. In some long-sequence settings, the dynamic mode even matches or exceeds the dense baseline, while mask-aware LoRA further improves long-horizon coherence. Code is available at https://github.com/Adamlong3/DynamicRad.

📄 PDF Abstract BibTeX arXiv:2604.20470

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Trainable Dynamic Mask Sparse Attention

2025-08-04 · Jingze Shi, Yifan Wu, Yiran Peng, Bingheng Wu 외 arxiv

The increasing demand for long-context modeling in large language models (LLMs) is bottlenecked by the quadratic complexity of the standard self-attention mechanism. The community has proposed sparse attention to mitigat…

BiSparse-AAS: Bilinear Sparse Attention and Adaptive Spans Framework for Scalable and Efficient Text Summarization

2025-10-31 · Desta Haileselassie Hagos, Legand L. Burge, Anietie Andy, Anis Yazidi 외 arxiv

Transformer-based architectures have advanced text summarization, yet their quadratic complexity limits scalability on long documents. This paper introduces BiSparse-AAS (Bilinear Sparse Attention with Adaptive Spans), a…

Text Summarization

SampleAttention: Near-Lossless Acceleration of Long Context LLM Inference with Adaptive Structured Sparse Attention

2024-06-17 · Qianchao Zhu, Jiangfei Duan, Chang Chen, Siran Liu 외

Large language models (LLMs) now support extremely long context windows, but the quadratic complexity of vanilla attention results in significantly long Time-to-First-Token (TTFT) latency. Existing approaches to address …

ReSSFormer: A Recursive Sparse Structured Transformer for Scalable and Long-Context Reasoning

2025-10-02 · Haochen You, Baojing Liu arxiv

While Transformer architectures have demonstrated impressive scalability across domains, they continue to face challenges in long-context reasoning, computational efficiency, and structural generalization - largely due t…

Computational Efficiency

Twilight: Adaptive Attention Sparsity with Hierarchical Top-$p$ Pruning

2025-02-04 · Chaofan Lin, Jiaming Tang, Shuo Yang, Hanshuo Wang 외

Leveraging attention sparsity to accelerate long-context large language models (LLMs) has been a hot research topic. However, current algorithms such as sparse attention or key-value (KV) cache compression tend to use a …