这项研究揭示了视频模型理解物理规律的能力差异，对从事视频理解、物理推理或自监督学习的开发者有直接参考价值，建议关注V-JEPA在时间动态建模上的优势。
该研究通过冻结特征探测方法，在IntPhys2和MVP基准上评估了三种视频基础模型（V-JEPA、VideoMAE、LTX-Video）对直觉物理信息的编码能力。V-JEPA在整体表现上最强，尤其在使用建模时间动态的探测时；VideoMAE表现接近，而LTX-Video虽较弱但仍能提取有效信号。分层分析显示，物理相关信息在早期层最弱，在中后期层最易获取；打乱帧顺序会显著降低性能，尤其在MVP上。结果表明，直觉物理知识在预训练视频表示中可靠涌现，但其可获取性依赖于预训练范式、表示深度和读出机制。
Do Video Foundation Models Understand Intuitive Physics? A Layerwise Probing Analysis
We study whether pretrained video foundation models encode intuitive-physics information in their frozen representations, and how this information varies across model families, layers, and probe types. Using frozen-feature probing on IntPhys2 and Minimal Video Pairs (MVP), we compare predictive joint-embedding models (V-JEPA), masked reconstruction models (VideoMAE), and a diffusion-based video generator (LTX-Video). V-JEPA achieves the strongest overall results across benchmarks, especially with probes that model temporal dynamics, while VideoMAE remains competitive and LTX-Video recovers weaker but non-trivial signal. Layerwise analyses show that physics-relevant information is weakest in early layers and becomes most accessible at intermediate-to-late depth, and temporal controls show that disrupting frame order substantially reduces performance, especially on MVP. Together, these results suggest that intuitive-physics knowledge emerges reliably in pretrained video representations, but its accessibility depends strongly on pretraining paradigm, representational depth, and readout mechanism.