沒有通道被遺忘:透過殘差閘理解擴散模型加速
No Channels Left Behind: Understanding Diffusion Acceleration through Residual Gates
研究發現,REPA 等擴散 Transformer 加速方法會重組 AdaLN-Zero 殘差閘 α,而較廣泛的閘參與度與較佳生成品質相關。研究提出無需教師模型的徑向目標,透過特徵範數正則化促進廣泛參與;在 ImageNet 256times256 的 SiT-B/2 實驗中,該方法的 FID 優於 dispersive loss。
Published on Oct 29
Authors:
,
,
,
Abstract
Training acceleration in diffusion transformers is driven by reorganization of AdaLN-Zero residual gates, and a simple radial objective promoting broader gate participation improves generation quality without auxiliary teachers.
Training acceleration methods for diffusion transformers, such as REPA and dispersive regularization, are commonly explained by additional supervision, spatial-structure priors, or improved feature geometry. We revisit this view and uncover a striking internal change: the AdaLN-Zero residual gate α, which directly controls how conditioned updates enter the residual stream, is substantially reorganized by these methods. This observation motivates our hypothesis that acceleration arises, at least in part, from reshaping the residual gating pathway rather than solely from the feature structures explicitly imposed by auxiliary objectives. Supporting this view, accelerated models are markedly more sensitive to perturbations of α, and their gate geometry reflects semantic relationships among classes. We further introduce the normalized participation ratio (nPR) to characterize the channel-wise organization of the gates and find that, across REPA teachers, projector architectures, and radial-objective targets, broader gate participation accompanies better generation quality. Guided by this insight, we propose a simple teacher-free radial objective that promotes broader gate participation through feature-norm regularization, without pairwise feature comparisons. On ImageNet 256times256 with SiT-B/2, our method improves FID over dispersive loss while increasing mean gate participation over timesteps and classes, further supporting the link between gate organization and training acceleration. Our results identify residual-gate reorganization as a measurable axis of diffusion training acceleration and suggest a simple direction for further acceleration.
View arXiv page View PDF Add to collection
Get this paper in your agent:
hf papers read 2610.40914
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash
Models citing this paper 0
No model linking this paper
Cite arxiv.org/abs/2610.40914 in a model README.md to link it from this page.
Datasets citing this paper 0
No dataset linking this paper
Cite arxiv.org/abs/2610.40914 in a dataset README.md to link it from this page.
Spaces citing this paper 0
No Space linking this paper
Cite arxiv.org/abs/2610.40914 in a Space README.md to link it from this page.
Collections including this paper 0
No Collection including this paper
Add this paper to a collection to link it from this page.
來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co