{
  "id": 714,
  "url": "https://arxiv.org/abs/2606.23019v1",
  "title": "ScalingAttention: Discovering Intrinsic Sparse Attention Topology for Video Diffusion Transformers",
  "summary": "While Diffusion Transformers (DiTs) have revolutionized high-fidelity video generation, their reliance on 3D full attention creates a quadratic computational bottleneck. Existing sparse methods face a dilemma: dynamic pruning suffers from prohibitive runtime overhead and memory fragmentation, while static heuristics fail to capture fine-grained dependencies. In this work, we propose ScalingAttention, a training-free framework grounded in a key inductive bias: while individual activations are inp",
  "authors": "Ruiliang Zhou, Xuecheng Wu, Kang He, Guangyun Han, Bin Liu, Qinqin Chen et al.",
  "category": "research",
  "topics": "bias-fairness",
  "orgs": null,
  "regions": null,
  "published_at": "2026-06-22T08:32:07.000Z",
  "fetched_at": "2026-07-14T14:14:46.033Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/714",
  "original_url": "https://arxiv.org/abs/2606.23019v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}