# MSC: Multi-Scale Spatio-Temporal Causal Attention for Autoregressive Video Diffusion

arXiv: https://arxiv.org/abs/2412.09828 · HTML: https://mlsanity.com/paper/2412.09828/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Placed in Video generation, one of 14 directions in Diffusion models.
- Builds on 9 of its 32 references that are in the map.
- First posted on arXiv 13 Dec 2024.

## Placed in

- Diffusion models / [Video generation](https://mlsanity.com/diffusion/video-generation/)

## Builds on (most cited first, 9 of 9)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [CogVideoX: Text-to-Video Diffusion Models with An Expert Transformer](https://mlsanity.com/paper/2408.06072/) (2024)
- [Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Prediction](https://mlsanity.com/paper/2404.02905/) (2024)
- [Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusion](https://mlsanity.com/paper/2407.01392/) (2024)
- [Movie Gen: A Cast of Media Foundation Models](https://mlsanity.com/paper/2410.13720/) (2024)
- [Language Model Beats Diffusion -- Tokenizer is Key to Visual Generation](https://mlsanity.com/paper/2310.05737/) (2023)
- [Open-Sora Plan: Open-Source Large Video Generation Model](https://mlsanity.com/paper/2412.00131/) (2024)
- [On the Importance of Noise Scheduling for Diffusion Models](https://mlsanity.com/paper/2301.10972/) (2023)

## Built on by (most cited first, 2 of 2)

- [Efficient Video Diffusion Models: Advancements and Challenges](https://mlsanity.com/paper/2604.15911/) (2026)
- [AvatarForcing: One-Step Streaming Talking Avatars via Local-Future Sliding-Window Denoising](https://mlsanity.com/paper/2603.14331/) (2026)

## Abstract (arXiv)

Diffusion transformers enable flexible generative modeling for video. However, it is still technically challenging and computationally expensive to generate high-resolution videos with rich semantics and complex motion. Similar to languages, video data are also auto-regressive by nature, so it is counter-intuitive to use attention mechanism with bi-directional dependency in the model. Here we propose a Multi-Scale Causal (MSC) framework to address these problems. Specifically, we introduce multiple resolutions in the spatial dimension and high-low frequencies in the temporal dimension to realize efficient attention calculation. Furthermore, attention blocks on multiple scales are combined in a controlled way to allow causal conditioning on noisy image frames for diffusion training, based on the idea that noise destroys information at different rates on different resolutions. We theoretically show that our approach can greatly reduce the computational complexity and enhance the efficiency of training. The causal attention diffusion framework can also be used for auto-regressive long video generation, without violating the natural order of frame sequences.