# LiteAttention: A Temporal Sparse Attention for Diffusion Transformers

arXiv: https://arxiv.org/abs/2511.11062 · HTML: https://mlsanity.com/paper/2511.11062/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- Placed in Video generation, one of 14 directions in Diffusion models.
- Builds on 5 of its 27 references that are in the map.
- First posted on arXiv 14 Nov 2025.

## Placed in

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

## Builds on (most cited first, 5 of 5)

- [Wan: Open and Advanced Large-Scale Video Generative Models](https://mlsanity.com/paper/2503.20314/) (2025)
- [LTX-Video: Realtime Video Latent Diffusion](https://mlsanity.com/paper/2501.00103/) (2024)
- [Online normalizer calculation for softmax](https://mlsanity.com/paper/1805.02867/) (2018)
- [Sparse VideoGen2: Accelerate Video Generation with Sparse Attention via Semantic-Aware Permutation](https://mlsanity.com/paper/2505.18875/) (2025)
- [SparseD: Sparse Attention for Diffusion Language Models](https://mlsanity.com/paper/2509.24014/) (2025)

## Built on by (most cited first, 12 of 12)

- [PISA: Piecewise Sparse Attention Is Wiser for Efficient Diffusion Transformers](https://mlsanity.com/paper/2602.01077/) (2026)
- [Mixture of Distributions Matters: Dynamic Sparse Attention for Efficient Video Diffusion Transformers](https://mlsanity.com/paper/2601.11641/) (2026)
- [Efficient Video Diffusion Models: Advancements and Challenges](https://mlsanity.com/paper/2604.15911/) (2026)
- [RoPeSLR: 3D RoPE-driven Sparse-LowRank Attention for Efficient Diffusion Transformers](https://mlsanity.com/paper/2605.20659/) (2026)
- [MobileWan: Closing the Quality Gap for Mobile Video Diffusion](https://mlsanity.com/paper/2607.06173/) (2026)
- [Accelerating Text-to-Video Generation with Calibrated Sparse Attention](https://mlsanity.com/paper/2603.05503/) (2026)
- [HEART: Exploiting Head Heterogeneity in Sparse Attention for Video Diffusion](https://mlsanity.com/paper/2605.14513/) (2026)
- [Chorus II: Cross-Request Sparsity Reuse for Efficient Image-to-Video Generation](https://mlsanity.com/paper/2606.25040/) (2026)
- [SCOPE: Subspace Clustering with Online Per-Head Top-K Estimation for Sparse Video Attention](https://mlsanity.com/paper/2608.12780/) (2026)
- [Where Compute Matters: Heterogeneous Attention for Efficient Video Diffusion](https://mlsanity.com/paper/2609.31050/) (2026)
- [iCATS: Fast Video Generation via Interaction-Aware Sparse Attention and Timestep-Adaptive Sparsity](https://mlsanity.com/paper/2610.11302/) (2026)
- [Sparse Attention for Video Generation Acceleration via Growing Sparsity and Reduced Search](https://mlsanity.com/paper/2610.21129/) (2026)

## Abstract (arXiv)

Diffusion Transformers, particularly for video generation, achieve remarkable quality but suffer from quadratic attention complexity, leading to prohibitive latency. Existing acceleration methods face a fundamental trade-off: dynamically estimating sparse attention patterns at each denoising step incurs high computational overhead and estimation errors, while static sparsity patterns remain fixed and often suboptimal throughout denoising. We identify a key structural property of diffusion attention, namely, its sparsity patterns exhibit strong temporal coherence across denoising steps. Tiles deemed non-essential at step t typically remain so at step t+δ. Leveraging this observation, we introduce LiteAttention, a method that exploits temporal coherence to enable evolutionary computation skips across the denoising sequence. By marking non-essential tiles early and propagating skip decisions forward, LiteAttention eliminates redundant attention computations without repeated profiling overheads, combining the adaptivity of dynamic methods with the efficiency of static ones. We implement a highly optimized LiteAttention kernel on top of FlashAttention and demonstrate substantial speedups on production video diffusion models, with no degradation in quality. The code and implementation details will be publicly released.