视频扩散模型中的因果性可与去噪分离
Causality in Video Diffusers is Separable from Denoising
把视频扩散模型中的时序因果推理与多步去噪解耦：用因果 Transformer 编码器每帧只做一次时序建模，再用轻量扩散解码器做逐帧多步渲染，可直接迁移到自回归视频扩散的加速与架构改造。
原文依据 we introduce Separable Causal Diffusion (SCD), a new architecture that explicitly decouples once-per-frame temporal reasoning
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论因果性——即组件之间时序的、单向的因果关系——是许多复杂生成过程的基础，包括视频、语言和机器人轨迹。当前的因果扩散模型将时序推理与迭代去噪纠缠在一起，在每一层、每个去噪步骤、整个上下文上施加因果注意力。本文表明，这些模型中的因果计算可以与多步去噪过程分离。通过对自回归视频扩散模型进行系统性探测，我们揭示了两条关键规律：(1) 浅层模块在不同去噪步骤间产生高度相似的特征，说明沿扩散轨迹存在冗余计算；(2) 更深的模块表现出稀疏的跨帧注意力，主要执行帧内渲染。受这些发现启发，我们提出可分离因果扩散（Separable Causal Diffusion, SCD），这是一种新架构，通过因果 Transformer 编码器将每帧一次的时序推理显式解耦，并通过轻量扩散解码器完成多步逐帧渲染。在合成与真实基准上的预训练与后训练任务中开展的大量实验表明，CSD 在显著提升吞吐量并降低延迟的同时，生成质量与强因果扩散基线相当或更优。
英文原文摘要 1345 字符
Causality--referring to temporal, uni-directional cause-effect relationships between components--underlies many complex generative processes, including videos, language, and robot trajectories.Current causal diffusion models entangle temporal reasoning with iterative denoising, applying causal attention across all layers, at every denoising step, and over the entire context.In this paper, we show that the causal computation in these models is separable from the multi-step denoising process.Through systematic probing of autoregressive video diffusers, we uncover two key regularities:(1) early blocks produce highly similar features across denoising steps, indicating redundant computation along the diffusion trajectory; and(2) deeper blocks exhibit sparse cross-frame attention and primarily perform intra-frame rendering.Motivated by these findings, we introduce Separable Causal Diffusion (SCD), a new architecture that explicitly decouples once-per-frame temporal reasoning, via a causal transformer encoder, from multi-step frame-wise rendering, via a lightweight diffusion decoder.Extensive experiments on both pretraining and post-training tasks across synthetic and real benchmarks show that CSD significantly improves throughput and latency while matching or surpassing the generation quality of strong causal diffusion baselines.