少量通道绘制整体画面：揭示扩散变换器中的大规模激活
Few Channels Draw The Whole Picture: Revealing Massive Activations in Diffusion Transformers
AI总结：
研究扩散变换器中少量关键通道对图像生成的决定性作用，揭示其在功能、空间组织和可迁移性上的核心贡献。
摘要：
扩散变换器（DiTs）及相关流架构已成为最强的文本到图像生成器，但提示如何塑造图像语义的内部机制仍不明确。本文研究大规模激活：一小部分隐藏状态通道，其响应显著大于其余部分。研究显示，尽管这些通道稀疏，但它们在三个互补意义上有效绘制整体画面：首先，它们功能关键，控制破坏测试显示删除大规模通道导致生成质量骤降，而破坏低统计通道影响微弱；其次，它们空间组织，限制图像流令牌到大规模通道并聚类可得到与主体和显著区域一致的结构化空间代码；第三，它们可迁移，将大规模激活从一个提示条件轨迹转移到另一个可使最终图像向源提示偏移同时保留目标内容，产生局部语义插值而非无结构像素混合。该特性在文本条件和图像条件语义传输中应用，通过大规模激活实现提示插值和主体驱动生成，无需额外训练。这些结果将大规模激活从激活异常重新定义为稀疏提示条件载波子空间，组织并控制现代DiT模型中的语义信息。
英文摘要：
Diffusion Transformers (DiTs) and related flow-based architectures are now among the strongest text-to-image generators, yet the internal mechanisms through which prompts shape image semantics remain poorly understood. In this work, we study massive activations: a small subset of hidden-state channels whose responses are consistently much larger than the rest. We show that, despite their sparsity, these few channels effectively draw the whole picture, in three complementary senses. First, they are functionally critical: a controlled disruption probe that zeroes the massive channels causes a sharp collapse in generation quality, while disrupting an equally-sized set of low-statistic channels has marginal effect. Second, they are spatially organized: restricting image-stream tokens to massive channels and clustering them yields coherent partitions that closely align with the main subject and salient regions, exposing a structured spatial code hidden inside an apparently outlier-like subspace. Third, they are transferable: transporting massive activations from one prompt-conditioned trajectory into another, shifts the final image toward the source prompt while preserving substantial content from the target, producing localized semantic interpolation rather than unstructured pixel blending. We exploit this property in two use cases: text-conditioned and image-conditioned semantic transport, where massive activations transport enables prompt interpolation and subject-driven generation without any additional training. Together, these results recast massive activations not as activation anomalies, but as a sparse prompt-conditioned carrier subspace that organizes and controls semantic information in modern DiT models.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
