发表机构
RIKEN iTHEMS; RIKEN AIP; South China University of Technology; Columbia University（理化学研究所 iTHEMS; 理化学研究所 AIP; 华南理工大学; 哥伦比亚大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究针对视频生成模型是否遵循物理不可逆性的问题，提出含进展与静态率的两部分协议，发现模型存在不可逆属性欠发展问题，并验证解耦属性潜空间的单调性构建可消除可操纵的读出引导。
AI 中文摘要
许多物理属性具有不可逆性：冰会融化但不会重新冻结，纸张会烧焦但不会恢复未燃烧状态。视频生成模型是否遵循这一特性？我们表明该问题难以测量，而可可靠测量的是「发展（development）」而非「逆转（reversal）」。局部逆转度量存在空退化问题：纯噪声下每片段违规率得分为0.50，经方差归一化的逆转残差处于噪声上限。通过空测试的是由两部分组成的协议：进展（progress，即定向属性相关性）和静态率（stasis rate）。在该协议下，生成视频与真实素材可清晰区分，且该差异已通过人类验证。针对7个文本到视频模型，真实参考素材的进展为ρ=+0.40、静态率为35%，而所有生成器的进展接近零、静态率为92%-100%；9名标注者的评分显示真实素材远高于生成素材（0-4分制下为2.75 vs. 0.99）。可靠发现是「欠发展（under-development）」：生成器几乎不会推进不可逆属性，而非逆转它们。作为补充机制，我们表明事后读出引导（post-hoc readout guidance）可被操纵，而通过在解耦属性潜空间中构建来强制单调性（monotonicity）可消除可操纵的读出，该结论在受控和半合成设置中得到验证。
英文摘要
Many physical attributes are \emph{irreversible}: ice melts but does not re-freeze, paper chars but does not un-burn. Do video generators respect this? We show the question is hard to measure, and that what can be measured reliably is \emph{development} rather than reversal. Metrics of local reversal are null-degenerate: a per-clip violation rate scores $0.50$ on pure noise, and a variance-normalized reversal residual sits at its noise ceiling. What survives null-testing is a two-part protocol: progress (a directional attribute correlation) and a stasis rate. Under this protocol, generated video separates cleanly from real footage, and the gap is human-validated. Across seven text-to-video models, real reference footage advances ($ρ{=}{+}0.40$, $35\%$ static) while every generator shows near-zero progress and $92$--$100\%$ stasis; nine annotators rate real footage far above generated ($2.75$ vs.\ $0.99$ on a $0$--$4$ scale). The reliable finding is \emph{under-development}: generators barely advance irreversible attributes rather than reversing them. As a complementary mechanism, we show that post-hoc readout guidance is gameable, whereas enforcing monotonicity by construction in a disentangled attribute latent removes the gameable readout, validated in controlled and semi-synthetic settings.
