mlsanityData to 7 Oct 2026

Causal Forcing: Autoregressive Diffusion Distillation Done Right for High-Quality Real-Time Interactive Video Generation

arXiv 2602.02214 ↗First posted 2 Feb 2026cs.CVcited by 159

On the map: Diffusion models → Video generation

Abstract (arXiv)

To achieve real-time interactive video generation, current methods distill pretrained bidirectional video diffusion models into few-step autoregressive (AR) models, facing an architectural gap when full attention is replaced by causal attention. However, existing approaches do not bridge this gap theoretically. They initialize the AR student via ODE distillation, which requires frame-level injectivity, where each noisy frame must map to a unique clean frame under the PF-ODE of an AR teacher. Distilling an AR student from a bidirectional teacher violates this condition, preventing recovery of the teacher's flow map and instead inducing a conditional-expectation solution, which degrades performance. To address this issue, we propose Causal Forcing, which uses an autoregressive teacher for ODE initialization to bridge the architectural gap, and then applies the same DMD procedure as in Self Forcing. Empirical results show that our method outperforms all baselines across all metrics, surpassing the SOTA Self Forcing by 19.3% in Dynamic Degree, 8.7% in VisionReward, and 16.7% in Instruction Following. Project page: hrefhttps://thu-ml.github.io/CausalForcing.github.io/https://thu-ml.github.io/CausalForcing.github.io/; the code: hrefhttps://github.com/thu-ml/Causal-Forcinghttps://github.com/thu-ml/Causal-Forcing.

Key facts. Cited by 159 papers in the map (data through 7 Oct 2026). Placed in Video generation, one of 14 directions in Diffusion models. Builds on 76 of its 83 references that are in the map. First posted on arXiv 2 Feb 2026.

Builds on

76 of 83 references.

Closest ancestors · 29

Same direction, newest first

  1. 2025Yume-1.5: A Text-Controlled Interactive World Generation Model26 Dec
  2. 2025Knot Forcing: Taming Autoregressive Video Diffusion Models for Real-time Infinite Interactive Portrait Animation25 Dec
  3. 2025End-to-End Training for Autoregressive Video Diffusion via Self-Resampling17 Dec
  4. 2025WorldPlay: Towards Long-Term Geometric Consistency for Real-Time Interactive World Modeling16 Dec
  5. 2025Deep Forcing: Training-Free Long Video Generation with Deep Sink and Participative Compression4 Dec
  6. 2025Infinity-RoPE: Action-Controllable Infinite Video Generation Emerges From Autoregressive Self-Rollout25 Nov
  7. 2025UltraViCo: Breaking Extrapolation Limits in Video Diffusion Transformers25 Nov
  8. 2025Self-Forcing++: Towards Minute-Scale High-Quality Video Generation2 Oct
  9. 2025Rolling Forcing: Autoregressive Long Video Diffusion in Real Time29 Sep
  10. 2025Radial Attention: O(nlog n) Sparse Attention with Energy Decay for Long Video Generation24 Jun
  11. 2025Self Forcing: Bridging the Train-Test Gap in Autoregressive Video Diffusion9 Jun
  12. 2025SkyReels-V2: Infinite-length Film Generative Model17 Apr
17 more
  1. 2025RIFLEx: A Free Lunch for Length Extrapolation in Video Diffusion Transformers21 Feb
  2. 2025Sparse VideoGen: Accelerating Video Diffusion Transformers with Spatial-Temporal Sparsity3 Feb
  3. 2024LTX-Video: Realtime Video Latent Diffusion30 Dec
  4. 2024Autoregressive Video Generation without Vector Quantization18 Dec
  5. 2024From Slow Bidirectional to Fast Autoregressive Video Diffusion Models10 Dec
  6. 2024Pyramidal Flow Matching for Efficient Video Generative Modeling8 Oct
  7. 2024CogVideoX: Text-to-Video Diffusion Models with An Expert Transformer12 Aug
  8. 2024Identifying and Solving Conditional Image Leakage in Image-to-Video Diffusion Model22 Jun
  9. 2024Vidu: a Highly Consistent, Dynamic and Skilled Text-to-Video Generator with Diffusion Models7 May
  10. 2024VidProM: A Million-scale Real Prompt-Gallery Dataset for Text-to-Video Diffusion Models10 Mar
  11. 2023Photorealistic Video Generation with Diffusion Models11 Dec
  12. 2023Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasets25 Nov
  13. 2023VideoCrafter1: Open Diffusion Models for High-Quality Video Generation30 Oct
  14. 2023DynamiCrafter: Animating Open-domain Images with Video Diffusion Priors18 Oct
  15. 2023Align your Latents: High-Resolution Video Synthesis with Latent Diffusion Models18 Apr
  16. 2022Latent Video Diffusion Models for High-Fidelity Long Video Generation23 Nov
  17. 2022Imagen Video: High Definition Video Generation with Diffusion Models5 Oct

Other references · 47

  1. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  2. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 8,964
  3. 2022Scalable Diffusion Models with Transformerscited by 5,931
  4. 2022Flow Matching for Generative Modelingcited by 5,426
  5. 2022Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flowcited by 3,550
  6. 2025Wan: Open and Advanced Large-Scale Video Generative Modelscited by 2,536
  7. 2023VBench: Comprehensive Benchmark Suite for Video Generative Modelscited by 1,755
  8. 2022Make-A-Video: Text-to-Video Generation without Text-Video Datacited by 1,725
  9. 2024HunyuanVideo: A Systematic Framework For Large Video Generative Modelscited by 1,597
  10. 2023ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with Variational Score Distillationcited by 1,207
  11. 2023One-step Diffusion with Distribution Matching Distillationcited by 978
  12. 2022CogVideo: Large-scale Pretraining for Text-to-Video Generation via Transformerscited by 973
35 more
  1. 2023Latent Consistency Models: Synthesizing High-Resolution Images with Few-Step Inferencecited by 826
  2. 2021VideoGPT: Video Generation using VQ-VAE and Transformerscited by 651
  3. 2024Open-Sora: Democratizing Efficient Video Production for Allcited by 637
  4. 2024Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusioncited by 634
  5. 2024Movie Gen: A Cast of Media Foundation Modelscited by 614
  6. 2025Mean Flows for One-step Generative Modelingcited by 524
  7. 2022All are Worth Words: A ViT Backbone for Diffusion Modelscited by 498
  8. 2023VideoPoet: A Large Language Model for Zero-Shot Video Generationcited by 479
  9. 2023Improved Techniques for Training Consistency Modelscited by 430
  10. 2023Understanding Diffusion Objectives as the ELBO with Simple Data Augmentationcited by 293
  11. 2024Open-Sora Plan: Open-Source Large Video Generation Modelcited by 286
  12. 2024Simplifying, Stabilizing and Scaling Continuous-Time Consistency Modelscited by 276
  13. 2025MAGI-1: Autoregressive Video Generation at Scalecited by 262
  14. 2021GODIVA: Generating Open-DomaIn Videos from nAtural Descriptionscited by 254
  15. 2021N"UWA: Visual Synthesis Pre-training for Neural visUal World creAtioncited by 253
  16. 2025LongLive: Real-time Interactive Long Video Generationcited by 237
  17. 2023Diff-Instruct: A Universal Approach for Transferring Knowledge From Pre-trained Diffusion Modelscited by 232
  18. 2019Scaling Autoregressive Video Modelscited by 189
  19. 2022EGSDE: Unpaired Image-to-Image Translation via Energy-Guided Stochastic Differential Equationscited by 172
  20. 2025History-Guided Video Diffusioncited by 165
  21. 2025Step-Video-T2V Technical Report: The Practice, Challenges, and Future of Video Foundation Modelcited by 163
  22. 2024VisionReward: Fine-Grained Multi-Dimensional Human Preference Learning for Image and Video Generationcited by 148
  23. 2024Consistency Models Made Easycited by 140
  24. 2025RELIC: Interactive Video World Model with Long-Horizon Memorycited by 89
  25. 2025Autoregressive Adversarial Post-Training for Real-Time Interactive Video Generationcited by 84
  26. 2025Large Scale Diffusion Distillation via Score-Regularized Continuous-Time Consistencycited by 81
  27. 2025MotionStream: Real-Time Video Generation with Interactive Motion Controlscited by 71
  28. 2025Hunyuan-GameCraft-2: Instruction-following Interactive Game World Modelcited by 58
  29. 2025Pack and Force Your Memory: Long-form and Consistent Video Generationcited by 26
  30. 2025StreamAvatar: Streaming Diffusion Models for Real-Time Interactive Human Avatarscited by 24
  31. 2023ControlVideo: Conditional Control for One-shot Text-driven Video Editing and Beyondcited by 22
  32. 2026Avatar Forcing: Real-Time Interactive Head Avatar Generation for Natural Conversationcited by 17
  33. 2025Towards One-step Causal Video Generation via Adversarial Self-Distillationcited by 15
  34. 2025Vidarc: Embodied Video Diffusion Model for Closed-loop Controlcited by 15
  35. 2025UltraImage: Rethinking Resolution Extrapolation in Image Diffusion Transformerscited by 10

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 159.

159 papers.

Sort:

By area:

  1. 2026Matrix-Game 3.0: Real-Time and Streaming Interactive World Model with Long-Horizon Memorycited by 86
  2. 2026Helios: Real Real-Time Long Video Generation Modelcited by 56
  3. 2026WoVR: World Models as Reliable Simulators for Post-Training VLA Policies with RLcited by 46
  4. 2026Causal Forcing++: Scalable Few-Step Autoregressive Diffusion Distillation for Real-Time Interactive Video Generationcited by 44
  5. 2026Infinite Worlds with Versatile Interactionscited by 37
  6. 2026DreamX-World 1.0: A General-Purpose Interactive World Modelcited by 33
  7. 2026Rolling Sink: Bridging Limited-Horizon Training and Open-Ended Testing in Autoregressive Video Diffusioncited by 28
  8. 2026ABot-M0.5: Unified Mobility-and-Manipulation World Action Modelcited by 28
  9. 2026LongLive-2.0: An NVFP4 Parallel Infrastructure for Long Video Generationcited by 25
  10. 2026MosaicMem: Hybrid Spatial Memory for Controllable Video World Modelscited by 23
  11. 2026Causal-rCM: A Unified Teacher-Forcing and Self-Forcing Open Recipe for Autoregressive Diffusion Distillation in Streaming Video Generation and Interactive World Modelscited by 23
  12. 2026MemRoPE: Training-Free Infinite Video Generation via Evolving Memory Tokenscited by 20
38 more
  1. 2026Relax Forcing: Relaxed KV-Memory for Consistent Long Video Generationcited by 20
  2. 2026MultiWorld: Scalable Multi-Agent Multi-View Video World Modelscited by 19
  3. 2026minWM: A Full-Stack Open-Source Framework for Real-Time Interactive Video World Modelscited by 19
  4. 2026Self Gradient Forcing: Native Long Video Extrapolationcited by 16
  5. 2026MBench: A Comprehensive Benchmark on Memory Capability for Video World Modelscited by 15
  6. 2026LiveEdit: Towards Real-Time Diffusion-Based Streaming Video Editingcited by 15
  7. 2026WorldKV: Efficient World Memory with World Retrieval and Compressioncited by 14
  8. 2026HiAR: Efficient Autoregressive Long Video Generation via Hierarchical Denoisingcited by 13
  9. 2026One-Forcing: Towards Stable One-Step Autoregressive Video Generationcited by 13
  10. 2026ActWorld: From Explorable to Interactive World Model via Action-Aware Memorycited by 13
  11. 2026ABot-World-0: Infinite Interactive World Rollout on a Single Desktop GPUcited by 13
  12. 2026Wonder: Video World Model Done Bettercited by 13
  13. 2026Alaya-EVOKE: From Linear-Scaling Supervision to Endless Worldcited by 13
  14. 2026OmniForcing: Unleashing Real-time Joint Audio-Visual Generationcited by 12
  15. 2026LongLive-RAG: A General Retrieval-Augmented Framework for Long Video Generationcited by 12
  16. 2026OPSD-V: On-Policy Self-Distillation for Post-Training Few-Step Autoregressive Video Generatorscited by 12
  17. 2026Salt: Self-Consistent Distribution Matching with Cache-Aware Training for Fast Video Generationcited by 11
  18. 2026Forcing-KV: Hybrid KV Cache Compression for Efficient Autoregressive Video Diffusion Modelscited by 10
  19. 2026SANA-Streaming: Real-time Streaming Video Editing with Hybrid Diffusion Transformercited by 10
  20. 2026GigaWorld-1: A Roadmap to Build World Models for Robot Policy Evaluationcited by 10
  21. 2026SVG-EAR: Parameter-Free Linear Compensation for Sparse Video Generation via Error-aware Routingcited by 9
  22. 2026Code World Model: Coding Agent as World Braincited by 9
  23. 2026SolarWM: Open Data and Scalable Training for Long-Horizon Video World Modelscited by 9
  24. 2026CausalCine: Real-Time Autoregressive Generation for Multi-Shot Video Narrativescited by 8
  25. 2026MSAVBench: Towards Comprehensive and Reliable Evaluation of Multi-Shot Audio-Video Generationcited by 8
  26. 2026Long-Horizon Streaming Video Generation via Hybrid Attention with Decoupled Distillationcited by 7
  27. 2026TetherCache: Stabilizing Autoregressive Long-Form Video Generation with Gated Recall and Trusted Alignmentcited by 7
  28. 2026Matrix-Game 3.5: Enhancing Real-Time Streaming Interactive World Models with Patch Memorycited by 7
  29. 2026AvatarForcing: One-Step Streaming Talking Avatars via Local-Future Sliding-Window Denoisingcited by 6
  30. 2026Efficient Video Diffusion Models: Advancements and Challengescited by 6
  31. 2026RAVEN: Real-time Autoregressive Video Extrapolation with Consistency-model GRPOcited by 6
  32. 2026One Sentence, One Drama: Personalized Short-Form Drama Generation via Multi-Agent Systemscited by 6
  33. 2026Echo-Infinity: Learning Evolving Memory for Real-Time Infinite Video Generationcited by 6
  34. 2026Flash-VAED: Plug-and-Play VAE Decoders for Efficient Video Generationcited by 5
  35. 2026Video Generation Models as World Models: Efficient Paradigms, Architectures and Algorithmscited by 5
  36. 2026Human Cognition in Machines: A Unified Perspective of World Modelscited by 5
  37. 2026A Systematic Post-Train Framework for Video Generationcited by 5
  38. 2026Xiaomi Auto World Model: A Joint World Model Integrating Reconstruction and Generation for Autonomous Drivingcited by 5

50 most cited of 159 All 159 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2025Playing with Transformer at 30+ FPS via Next-Frame Diffusion
  2. 2025AR-Diffusion: Asynchronous Video Generation with Auto-Regressive Diffusion
  3. 2026Flex-Forcing: Towards a Unified Autoregressive and Bidirectional Video Diffusion Model
  4. 2026UniTemp: Unlocking Video Generation in Any Temporal Order via Bidirectional Distillation
  5. 2026SoulX-LiveAct: Towards Hour-Scale Real-Time Human Animation with Neighbor Forcing and ConvKV Memory
  6. 2026MV-Forcing: Long Multi-View Video Generation via 4D-Grounded Spatio-Temporal Self-Forcing
  7. 2026In-Context Forcing: Uncovering Context Effects in Autoregressive Video Diffusion
  8. 2026Light Forcing: Accelerating Autoregressive Video Diffusion via Sparse Attention