{"slug": "ab-attention-accurate-block-attention-for-accelerating-video-diffusion", "title": "AB-Attention: Accurate Block Attention for Accelerating Video Diffusion Transformers", "summary": "Researchers present AB-Attention, an accurate block attention method for accelerating video diffusion transformers, addressing the quadratic cost of 3D full attention that dominates inference. The work targets a fidelity problem in dynamic block-sparse attention, which reduces cost by computing exact attention only on key/value blocks selected from current activations but whose accuracy degrades under that approximation. The source provides no measured speedup, benchmark figure, or named author for the method.", "body_md": "Video diffusion transformers generate high-quality videos, but the quadratic cost of 3D full attention dominates their inference. Dynamic block-sparse attention reduces this cost by computing exact attention only on key/value blocks selected from current activations. However, its fidelity degrades u", "url": "https://wpnews.pro/news/ab-attention-accurate-block-attention-for-accelerating-video-diffusion", "canonical_source": "https://aiflash.com/news/142298/", "published_at": "2026-10-30 08:30:22+00:00", "updated_at": "2026-10-30 08:47:42.386389+00:00", "lang": "en", "topics": ["artificial-intelligence", "generative-ai", "machine-learning", "ai-research"], "entities": ["AB-Attention", "video diffusion transformers"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/ab-attention-accurate-block-attention-for-accelerating-video-diffusion", "markdown": "https://wpnews.pro/news/ab-attention-accurate-block-attention-for-accelerating-video-diffusion.md", "text": "https://wpnews.pro/news/ab-attention-accurate-block-attention-for-accelerating-video-diffusion.txt", "jsonld": "https://wpnews.pro/news/ab-attention-accurate-block-attention-for-accelerating-video-diffusion.jsonld"}}