🛰️ Daily AI Frontier
‹ back to 2026-08-04

Token Radius Attention for Efficient Video Generation

Research Efficiency & Systems

Ranking

Overall 67
Content 80
Popularity 36

Observed public metrics from 1 member.

Representative image for Token Radius Attention for Efficient Video Generation

Merged summary

TL;DR - A training-free sparse-attention method, Token Radius Attention (TRA), cuts the quadratic cost of 3D self-attention in Video Diffusion Transformers by allocating per-query compute budgets from attention entropy. It delivers ~1.6–2x inference speedup on production video models without retraining.

  • Core observation: retained attention density varies per query but correlates log-linearly with attention entropy, and dominant interactions form query-centered neighborhoods with token-dependent radii — unlike head/block-level methods that share one budget across queries.
  • TRA maps query entropy to an analytic token budget, then converts it into a temporally decayed radius, avoiding explicit key ranking/sorting overhead.
  • Systems-level optimizations: fused entropy extraction, warm-up reuse across steps, and block-sparse mask construction to keep overhead low.
  • Reported results: 9–19% of attention interactions retained and 1.56x–2.05x speedup across seven Wan2.1, Wan2.2, and HunyuanVideo T2V/I2V configurations with competitive quality; code released.

Sources (1)

Token Radius Attention for Efficient Video Generation

arXiv cs.CV Jiayu Chen, Zhikun Jiang, Maoliang Li, Jiayi Luo, Jiawei Yang, Zihao Zheng, Hengyi Zhang, Guojie Luo, Xiang Chen 2026-08-03 arXiv:2608.02504
Public signals Semantic Scholar citations 0 · Semantic Scholar influential citations 0
Providers: Hugging Face · N/A OpenAlex · N/A Publisher · N/A Semantic Scholar · Citations 0 · Influential citations 0 X · N/A Fetched 2026-08-28 14:28:56.809323 UTC

TL;DR - A training-free sparse-attention method, Token Radius Attention (TRA), cuts the quadratic cost of 3D self-attention in Video Diffusion Transformers by allocating per-query compute budgets from attention entropy. It delivers ~1.6–2x inference speedup on production video models without retraining.

  • Core observation: retained attention density varies per query but correlates log-linearly with attention entropy, and dominant interactions form query-centered neighborhoods with token-dependent radii — unlike head/block-level methods that share one budget across queries.
  • TRA maps query entropy to an analytic token budget, then converts it into a temporally decayed radius, avoiding explicit key ranking/sorting overhead.
  • Systems-level optimizations: fused entropy extraction, warm-up reuse across steps, and block-sparse mask construction to keep overhead low.
  • Reported results: 9–19% of attention interactions retained and 1.56x–2.05x speedup across seven Wan2.1, Wan2.2, and HunyuanVideo T2V/I2V configurations with competitive quality; code released.
item →