# Supplymentary Materials 
*NeurIPS 2025 submission 110* 

VORTA: Efficient Video Diffusion via Routing Sparse Attention


The files are formatted as follows:

```
<model_name>/                   # {HunyuanVideo, Wan-14B, bad_case}
├── <method_name>/              # {<model_name>, VORTA, VORTA+PAB/PCD, ARnR, STA, PAB, PCD}
│   ├── <video_prompt_0>.mp4    # The video names are same as the prompt in VBench
│   ├── ...
│   └── <video_prompt_4>.mp4
└── 
```


## Code
The code is available at anonymized [repository](https://anonymous.4open.science/r/VORTA-80C8/README.md).


## Technical appendix file
The technical appendix, provided as '110_VORTA_Appendix.pdf', contains supplementary details and additional results referenced in the main manuscript. It uses the same reference list to ensure consistency.


## Videos
The video results referenced in the main content and appendix of the paper are included in these supplementary materials for cross-reference. Detailed video specifications are provided in the appendix. 
