发表机构
Beijing University of Posts and Telecommunications（北京邮电大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究首次揭示视频扩散模型违反物理规律的根因在于RoPE导致的空间注意力过度衰减，并提出缩放RoPE频率的轻量级架构修改，有效增强生成视频的物理常识。
AI 中文摘要
尽管视觉质量令人印象深刻，最先进的视频扩散模型常常生成违反现实世界物理规律的内容。现有解决方案依赖外部先验或专门数据，而我们通过探索这些模型的内部机制来研究根本原因。具体而言，我们首次对文本到视频扩散模型的“运动规划”过程进行了可解释性研究，揭示了运动轨迹在早期去噪阶段是如何形成的。基于“先形状，后细节”的发现，我们将交叉注意力轨迹模式与因果头贡献相结合，识别出驱动运动规划的特定注意力头子集。此外，我们的自注意力分析表明，旋转位置嵌入（RoPE）会导致过度的空间注意力衰减。这导致早期候选区域过早锁定在物理上不合理的位置，抑制了相邻帧中的合理轨迹，并触发生成失败模式。为解决这一根本缺陷，我们提出了一种轻量级架构修改，在不同去噪步骤中缩放RoPE的频率。该策略减少了过度的注意力衰减，帮助模型探索更好的候选区域以建立连贯的物理运动。最后，免训练和基于训练的实验证实了我们的方法在增强生成视频的物理常识方面的有效性。
英文摘要
Despite impressive visual quality, state-of-the-art video diffusion models often generate content that violates real-world physical laws. While existing solutions rely on external priors or specialized data, we investigate the root cause by exploring the internal mechanisms of these models. Specifically, we present the first interpretability study on the ''motion planning'' process of text-to-video diffusion models, revealing how motion trajectories form during early denoising stages. Building upon the ''first shape, then details'' finding, we combine cross-attention trajectory patterns with causal head contributions to identify a specific subset of attention heads driving motion planning. Further, our self-attention analysis shows that Rotary Position Embedding (RoPE) induces excessive spatial attention decay. This causes early candidate regions to prematurely lock into physically implausible positions, suppressing reasonable trajectories in adjacent frames and triggering generation failure modes. To address this fundamental flaw, we propose a lightweight architectural modification that scales the frequency of RoPE across different denoising steps. This strategy reduces excessive attention decay, helping the model explore better candidate regions to establish coherent physical motion. Finally, training-free and training-based experiments confirm the effectiveness of our approach in enhancing the physical commonsense of generated videos.
