Skip to content
Hot eventLive

DDT-RFE:去除残差连接的解耦扩散Transformer

1 reports1 sources3 hr ago updated

Get the story

AI overview

研究者提出 DDT-RFE,针对 Decoupled Diffusion Transformer(DDT)编码器块的结构进行改进。该方法移除了 Self-Attention 与 MLP 周围的残差连接,并将输入 patch embedding 与中间及最终编码器特征融合,使解码器能够获得多层信息。论文发表于 arXiv(Machine Learning Theory),目前仅见方法提出,暂无实验结果或对比数据报道。

Generated from reports · updated 3 hr ago

Timeline

Follow the coverage from different angles.

Oct 7, 2026
  1. arXiv · Machine Learning Theory
    DDT-RFE:是否应跳过 Diffusion 模型中的残差连接

    研究者提出 DDT-RFE,移除 Decoupled Diffusion Transformer(DDT)编码器块中 Self-Attention 与 MLP 周围的残差连接,并将输入 patch embedding 与中间及最终编码器特征融合,使解码器获得多层信息。

Heat trend

Current heat 9·Comparable peak 10(Oct 7)·Comparable change over 24 hours –

02.557.510Oct7Oct7Oct7Oct7

The trend compares only the same participants observed continuously; its range may be smaller than the current heat count. Move or click on the chart to inspect hourly heat; use the left and right arrow keys to switch.