mlsanityData to 14 Oct 2026

Block Cascading: Training Free Acceleration of Block-Causal Video Models

arXiv 2511.20426 ↗First posted 25 Nov 2025cs.CV cs.AIcited by 6

Not on the map. Why?

Abstract (arXiv)

Block-causal video generation faces a stark speed-quality trade-off: small 1.3B models manage only 16 FPS while large 14B models crawl at 4.5 FPS, forcing users to choose between responsiveness and quality. Block Cascading significantly mitigates this trade-off through training-free parallelization. Our key insight: future video blocks do not need fully denoised current blocks to begin generation. By starting block generation with partially denoised context from predecessors, we transform sequential pipelines into parallel cascades where multiple blocks denoise simultaneously. With 5 GPUs exploiting temporal parallelism, we achieve ~2x acceleration across all model scales: 1.3B models accelerate from 16 to 30 FPS, 14B models from 4.5 to 12.5 FPS. Beyond inference speed, Block Cascading eliminates overhead from KV-recaching (of ~200ms) during context switches for interactive generation. Extensive evaluations validated against multiple block-causal pipelines demonstrate no significant loss in generation quality when switching from block-causal to Block Cascading pipelines for inference. Project Page: https://hmrishavbandy.github.io/block_cascading_page/

Key facts. Cited by 6 papers in the map (data through 14 Oct 2026). Builds on 54 of its 57 references that are in the map. First posted on arXiv 25 Nov 2025.

Builds on

54 of 57 references.

References · 54

  1. 2017Attention Is All You Needcited by 60,361
  2. 2020Language Models are Few-Shot Learnerscited by 32,419
  3. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  4. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  5. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 9,122
  6. 2024Qwen2.5 Technical Reportcited by 4,942
  7. 2024Scaling Rectified Flow Transformers for High-Resolution Image Synthesiscited by 4,181
  8. 2025Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilitiescited by 3,796
  9. 2022DreamFusion: Text-to-3D using 2D Diffusioncited by 3,025
  10. 2025Wan: Open and Advanced Large-Scale Video Generative Modelscited by 2,708
  11. 2024CogVideoX: Text-to-Video Diffusion Models with An Expert Transformercited by 2,261
  12. 2023Efficient Streaming Language Models with Attention Sinkscited by 2,186
38 more
  1. 2019Transformer-XL: Attentive Language Models Beyond a Fixed-Length Contextcited by 1,850
  2. 2023VBench: Comprehensive Benchmark Suite for Video Generative Modelscited by 1,822
  3. 2024HunyuanVideo: A Systematic Framework For Large Video Generative Modelscited by 1,660
  4. 2020Linformer: Self-Attention with Linear Complexitycited by 1,474
  5. 2023ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with Variational Score Distillationcited by 1,219
  6. 2023One-step Diffusion with Distribution Matching Distillationcited by 1,019
  7. 2024Improved Distribution Matching Distillation for Fast Image Synthesiscited by 684
  8. 2024Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusioncited by 681
  9. 2021VideoGPT: Video Generation using VQ-VAE and Transformerscited by 658
  10. 2023ModelScope Text-to-Video Technical Reportcited by 658
  11. 2024Movie Gen: A Cast of Media Foundation Modelscited by 644
  12. 2025Self Forcing: Bridging the Train-Test Gap in Autoregressive Video Diffusioncited by 625
  13. 2024KVQuant: Towards 10 Million Context Length LLM Inference with KV Cache Quantizationcited by 591
  14. 2022Efficiently Scaling Transformer Inferencecited by 539
  15. 2024FlashAttention-3: Fast and Accurate Attention with Asynchrony and Low-precisioncited by 519
  16. 2023Ring Attention with Blockwise Transformers for Near-Infinite Contextcited by 457
  17. 2024From Slow Bidirectional to Fast Autoregressive Video Diffusion Modelscited by 433
  18. 2021StyleGAN-V: A Continuous Video Generator with the Price, Image Quality and Perks of StyleGAN2cited by 301
  19. 2025MAGI-1: Autoregressive Video Generation at Scalecited by 276
  20. 2024SDXL-Lightning: Progressive Adversarial Diffusion Distillationcited by 266
  21. 2025LongLive: Real-time Interactive Long Video Generationcited by 265
  22. 2025Rolling Forcing: Autoregressive Long Video Diffusion in Real Timecited by 234
  23. 2025SkyReels-V2: Infinite-length Film Generative Modelcited by 229
  24. 2024Score identity Distillation: Exponentially Fast Distillation of Pretrained Diffusion Models for One-Step Generationcited by 214
  25. 2025Step-Video-T2V Technical Report: The Practice, Challenges, and Future of Video Foundation Modelcited by 168
  26. 2021A Good Image Generator Is What You Need for High-Resolution Video Synthesiscited by 163
  27. 2025Diffusion Adversarial Post-Training for One-Step Video Generationcited by 151
  28. 2025Frame Context Packing and Drift Prevention in Next-Frame-Prediction Video Diffusion Modelscited by 143
  29. 2025GameFactory: Creating New Games with Generative Interactive Videoscited by 130
  30. 2025Fast Video Generation with Sliding Tile Attentioncited by 128
  31. 2025VSA: Faster Video Diffusion with Trainable Sparse Attentioncited by 104
  32. 2025SANA-Video: Efficient Video Generation with Block Linear Diffusion Transformercited by 94
  33. 2025Autoregressive Adversarial Post-Training for Real-Time Interactive Video Generationcited by 90
  34. 2025SANA-Sprint: One-Step Diffusion with Continuous-Time Consistency Distillationcited by 87
  35. 2023ARTboldsymbol·V: Auto-Regressive Text-to-Video Generation with Diffusion Modelscited by 76
  36. 2025MotionStream: Real-Time Video Generation with Interactive Motion Controlscited by 76
  37. 2024SF-V: Single Forward Video Generation Modelcited by 40
  38. 2024ViD-GPT: Introducing GPT-style Autoregressive Generation in Video Diffusion Modelscited by 36

50 most cited of 54 All 54 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 6.

6 papers.

  1. 2026Continuous Latent Diffusion Language Modelcited by 14
  2. 2026Attention Sink in Transformers: A Survey on Utilization, Interpretation, and Mitigationcited by 11
  3. 2026Efficient Video Diffusion Models: Advancements and Challengescited by 7
  4. 2026X-Cache: Cross-Chunk Block Caching for Few-Step Autoregressive World Models Inferencecited by 7
  5. 2026When to Lock Attention: Training-Free KV Control in Video Diffusioncited by 0
  6. 2026UnStep: Training-Free Acceleration of Causal Video Diffusion with Fewer Steps Than Distillationcited by 0