Metraipolis › archive › dc_baefd7991bf8

SFA: A Statistical Physics-Inspired Diffusion Transformer for Efficient and High-Quality Image Generation — arXiv:2611.00027

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.00027
License (as archived)
CC0-1.0
Archived
2026-11-03T10:45:38.448Z
SHA-256 of the archived text
fab385f1f3ad0800632776961ec252fdb9c5b23c643e903e242cd3fd6bed7889
Stable URL of this page
https://metraipolis.com/documento/dc_baefd7991bf8

How to cite

SFA: A Statistical Physics-Inspired Diffusion Transformer for Efficient and High-Quality Image Generation — arXiv:2611.00027 — arXiv (en), https://arxiv.org/abs/2611.00027 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_baefd7991bf8 (Metraipolis archive dc_baefd7991bf8, sha256 fab385f1f3ad0800632776961ec252fdb9c5b23c643e903e242cd3fd6bed7889)
BibTeX
@misc{dcbaefd7991bf8,
  title = {SFA: A Statistical Physics-Inspired Diffusion Transformer for Efficient and High-Quality Image Generation — arXiv:2611.00027},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.00027},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_baefd7991bf8; sha256 fab385f1f3ad0800632776961ec252fdb9c5b23c643e903e242cd3fd6bed7889},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.00027. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Diffusion transformers have achieved remarkable success in image generation, but accelerating attention while preserving generation quality remains a major challenge at high resolutions. Inspired by statistical physics and the widespread occurrence of scale-free laws in nature, we introduce Scale-Free Attention (SFA), which preserves fine local interactions while summarizing distant regions at progressively coarser scales. For $N$ tokens, mean pooling, token-count correction, and a shared softmax over a scale-free spatial hierarchy give SFA $O(N\log N)$ complexity in both forward and backward passes. Our Triton implementation achieves $357.3\times$ forward and $265.6\times$ backward attention speedups over FlashAttention at 262,144 tokens. After 40 epochs of PixelFlow training from scratch on ImageNet at $256\times256$ resolution, SFA achieves the best generation quality among five attention mechanisms, including full attention. Fréchet inception distance evaluated on 50,000 generated samples (FID-50k) improves from 13.25 for full attention to 12.51, and the advantage over full attention persists through 80 epochs. SFA also achieves the best generation quality among the compared efficient-attention methods on Flickr-Faces-HQ (FFHQ). Beyond training from scratch, Stable Diffusion 3.5 Medium with attention replaced by SFA retains the generation quality of its full-attention counterpart under matched fine-tuning budgets. We obtain similar results when fine-tuning pretrained PixelFlow on ImageNet. SFA establishes scale-free structure as an effective design principle for efficient, high-quality diffusion transformers.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI); Graphics (cs.GR); Image and Video Processing (eess.IV).
Pagina del lavoro: https://arxiv.org/abs/2611.00027 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-03T10:39:50.363Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).