UnStep: Training-Free Acceleration of Causal Video Diffusion with Fewer Steps Than Distillation

arXiv:2609.32518 · cs.CV · Submitted 2026-09-26 · Read on arXiv

cs.CV

Submitted: 2026-09-26

Updated: 2026-09-26

Code: https://github.com/facebookresearch/UnStep

Project page: https://y-mansour.github.io/UnStep

Terminology

• Distillation — 34× in this paper · explained in 2 episodes
• Diffusion Model — 24× in this paper · explained in 8 episodes
• Consistency — 20× in this paper · explained in 3 episodes
• KV cache — 20× in this paper · explained in 3 episodes
• Transformer — 18× in this paper · explained in 2 episodes
• Diffusion Models — 17× in this paper · explained in 6 episodes
• Benchmark — 13× in this paper · explained in 2 episodes
• In International Conference on Learning Representations — 11× in this paper
• Sparse Attention — 10× in this paper · explained in 5 episodes
• Distribution Matching — 9× in this paper · explained in 2 episodes
• Computer Vision and Pattern Recognition — 8× in this paper
• Fidelity — 8× in this paper · explained in 2 episodes
• Decode — 7× in this paper · explained in 2 episodes
• Quantization — 7× in this paper · explained in 12 episodes
• Ablation — 6× in this paper · explained in 1 episode
• In International Conference on Machine Learning — 6× in this paper
• Sparsity — 5× in this paper · explained in 4 episodes
• Alignment — 4× in this paper · explained in 1 episode
• Generative Model — 3× in this paper · explained in 2 episodes
• Generative Models — 3× in this paper · explained in 2 episodes
• Adaptation — 2× in this paper · explained in 1 episode
• Aggregation — 2× in this paper · explained in 1 episode
• Benchmarking — 2× in this paper · explained in 3 episodes
• Embeddings — 2× in this paper · explained in 2 episodes
• Flow Matching — 2× in this paper · explained in 7 episodes
• 3D Positional Encoding — 1× in this paper · explained in 1 episode
• CVF International Conference on Computer Vision — 1× in this paper
• Communication Overhead — 1× in this paper · explained in 2 episodes
• Core — 1× in this paper · explained in 2 episodes
• Foundation Model — 1× in this paper · explained in 3 episodes
• In European Conference on Computer Vision — 1× in this paper
• Latent Diffusion Model — 1× in this paper · explained in 2 episodes
• Latent Diffusion Models — 1× in this paper · explained in 2 episodes
• Latent Space — 1× in this paper · explained in 7 episodes
• Optical Flow — 1× in this paper · explained in 2 episodes
• Pruning — 1× in this paper · explained in 4 episodes
• Self-distillation — 1× in this paper · explained in 6 episodes
• Transformer Models — 1× in this paper · explained in 2 episodes
• Verification — 1× in this paper · explained in 2 episodes
• World Model — 1× in this paper · explained in 7 episodes

Sources

• MonarchRT: Efficient Attention for Real-Time Video Generation
• LongLive-2.0: An NVFP4 Parallel Infrastructure for Long Video Generation
• One-Forcing: Towards Stable One-Step Autoregressive Video Generation
• AnyFlow: Any-Step Video Diffusion Model with On-Policy Flow Map Distillation
• Efficient Autoregressive Video Diffusion with Dummy Head
• Forcing-KV: Hybrid KV Cache Compression for Efficient Autoregressive Video Diffusion Models
• DistillAlign: Coordinating Mode Covering and Mode Seeking in Autoregressive Video Distillation
• Long-Horizon Streaming Video Generation via Hybrid Attention with Decoupled Distillation
• Ms. Forcing: Efficient Streaming Video Generation with Multi-Scale Patchification and Attention
• SDXL-Lightning: Progressive Adversarial Diffusion Distillation
• SlimDiff: Training-Free, Activation-Guided Hands-free Slimming of Diffusion Models
• HeadCast: Casting Attention Heads for Efficient Autoregressive Video Generation
• Wan: Open and Advanced Large-Scale Video Generative Models
• Sparse Forcing: Native Trainable Sparse Attention for Real-time Autoregressive Diffusion Video Generation
• In-Context Forcing: Uncovering Context Effects in Autoregressive Video Diffusion
• VideoSSM: Autoregressive Long Video Generation with Hybrid State-Space Memory
• Accelerating Video Generation Inference with Sequential-Parallel 3D Positional Encoding Using a Global Time Index
• Helios: Real Real-Time Long Video Generation Model
• Causal Forcing++: Scalable Few-Step Autoregressive Diffusion Distillation for Real-Time Interactive Video Generation
• Causal-rCM: A Unified Teacher-Forcing and Self-Forcing Open Recipe for Autoregressive Diffusion Distillation in Streaming Video Generation and Interactive World Models

Related papers

• Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
• AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
• TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
• A Survey on Efficient Vision-Language-Action Models