mlsanityData to 14 Oct 2026

One-Minute Video Generation with Test-Time Training

arXiv 2504.05298 ↗First posted 7 Apr 2025cs.CVcited by 101

On the map: State-space sequence models (S4, Mamba) → Video, motion and generative Mamba models

Abstract (arXiv)

Transformers today still struggle to generate one-minute videos because self-attention layers are inefficient for long context. Alternatives such as Mamba layers struggle with complex multi-scene stories because their hidden states are less expressive. We experiment with Test-Time Training (TTT) layers, whose hidden states themselves can be neural networks, therefore more expressive. Adding TTT layers into a pre-trained Transformer enables it to generate one-minute videos from text storyboards. For proof of concept, we curate a dataset based on Tom and Jerry cartoons. Compared to baselines such as Mamba~2, Gated DeltaNet, and sliding-window attention layers, TTT layers generate much more coherent videos that tell complex stories, leading by 34 Elo points in a human evaluation of 100 videos per method. Although promising, results still contain artifacts, likely due to the limited capability of the pre-trained 5B model. The efficiency of our implementation can also be improved. We have only experimented with one-minute videos due to resource constraints, but the approach can be extended to longer videos and more complex stories. Sample videos, code and annotations are available at: https://test-time-training.github.io/video-dit

Key facts. Cited by 101 papers in the map (data through 14 Oct 2026). Placed in Video, motion and generative Mamba models, one of 10 directions in State-space sequence models (S4, Mamba). Builds on 50 of its 57 references that are in the map. First posted on arXiv 7 Apr 2025.

Builds on

50 of 57 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024Scaling Diffusion Mamba with Bidirectional SSMs for Efficient Image and Video Generation24 May

Other references · 49

  1. 2020Denoising Diffusion Implicit Modelscited by 9,781
  2. 2022Scalable Diffusion Models with Transformerscited by 6,054
  3. 2022Classifier-Free Diffusion Guidancecited by 5,847
  4. 2022Generative Adversarial Networkscited by 5,627
  5. 2023Mamba: Linear-Time Sequence Modeling with Selective State Spacescited by 4,718
  6. 2022Flamingo: a Visual Language Model for Few-Shot Learningcited by 4,386
  7. 2022FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awarenesscited by 3,673
  8. 2019Analyzing and Improving the Image Quality of StyleGANcited by 3,588
  9. 2016Gaussian Error Linear Units (GELUs)cited by 3,502
  10. 2020Longformer: The Long-Document Transformercited by 3,413
  11. 2024CogVideoX: Text-to-Video Diffusion Models with An Expert Transformercited by 2,261
  12. 2020Transformers are RNNs: Fast Autoregressive Transformers with Linear Attentioncited by 2,022
37 more
  1. 2019Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelismcited by 1,868
  2. 2024HunyuanVideo: A Systematic Framework For Large Video Generative Modelscited by 1,660
  3. 2024Transformers are SSMs: Generalized Models and Efficient Algorithms Through Structured State Space Dualitycited by 1,336
  4. 2024Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preferencecited by 1,184
  5. 2022MaskGIT: Masked Generative Image Transformercited by 1,099
  6. 2022CogVideo: Large-scale Pretraining for Text-to-Video Generation via Transformerscited by 985
  7. 2021Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure Synthetic Datacited by 970
  8. 2024Movie Gen: A Cast of Media Foundation Modelscited by 644
  9. 2024Gated Delta Networks: Improving Mamba2 with Delta Rulecited by 532
  10. 2024FlashAttention-3: Fast and Accurate Attention with Asynchrony and Low-precisioncited by 519
  11. 2021Linear Transformers Are Secretly Fast Weight Programmerscited by 502
  12. 2022Phenaki: Variable Length Video Generation From Open Domain Textual Descriptioncited by 484
  13. 2024Parallelizing Linear Transformers with the Delta Rule over Sequence Lengthcited by 396
  14. 2022Latent Video Diffusion Models for High-Fidelity Long Video Generationcited by 388
  15. 2024Learning to (Learn at Test Time): RNNs with Expressive Hidden Statescited by 355
  16. 2024Titans: Learning to Memorize at Test Timecited by 328
  17. 2023Common Diffusion Noise Schedules and Sample Steps are Flawedcited by 310
  18. 2023Photorealistic Video Generation with Diffusion Modelscited by 306
  19. 2023Effective Long-Context Scaling of Foundation Modelscited by 303
  20. 2021StyleGAN-V: A Continuous Video Generator with the Price, Image Quality and Perks of StyleGAN2cited by 301
  21. 2024StreamingT2V: Consistent, Dynamic, and Extendable Long Video Generation from Textcited by 251
  22. 2024StoryDiffusion: Consistent Self-Attention for Long-Range Image and Video Generationcited by 247
  23. 2022Long Video Generation with Time-Agnostic VQGAN and Time-Sensitive Transformercited by 244
  24. 2023SEINE: Short-to-Long Video Diffusion Model for Generative Transition and Predictioncited by 233
  25. 2016Visual Storytellingcited by 176
  26. 2023NUWA-XL: Diffusion over Diffusion for eXtremely Long Video Generationcited by 171
  27. 2018StoryGAN: A Sequential Conditional GAN for Story Visualizationcited by 138
  28. 2021Going Beyond Linear Transformers with Recurrent Fast Weight Programmerscited by 89
  29. 2023Intelligent Grimm -- Open-ended Visual Storytelling via Latent Diffusion Modelscited by 77
  30. 2022Make-A-Story: Visual Memory Conditioned Consistent Story Generationcited by 76
  31. 2022StoryDALL-E: Adapting Pretrained Text-to-Image Transformers for Story Continuationcited by 74
  32. 2025Test-time regression: a unifying framework for designing sequence models with associative memorycited by 74
  33. 2022Synthesizing Coherent Story with Auto-Regressive Latent Diffusion Modelscited by 70
  34. 2018Imagine This! Scripts to Compositions to Videoscited by 54
  35. 2024LinGen: Towards High-Resolution Minute-Length Text-to-Video Generation with Linear Computational Complexitycited by 43
  36. 2024ThunderKittens: Simple, Fast, and Adorable AI Kernelscited by 36
  37. 2022Meta-Learning Fast Weight Language Modelscited by 21

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 54; 2026: 47.

101 papers.

Sort:

By area:

  1. 2025TTRL: Test-Time Reinforcement Learningcited by 275
  2. 2025LongLive: Real-time Interactive Long Video Generationcited by 265
  3. 2025Frame Context Packing and Drift Prevention in Next-Frame-Prediction Video Diffusion Modelscited by 143
  4. 2025Hunyuan-GameCraft: High-dynamic Interactive Game Video Generation with Hybrid History Conditioncited by 98
  5. 2025RELIC: Interactive Video World Model with Long-Horizon Memorycited by 96
  6. 2025Mixture of Contexts for Long Video Generationcited by 95
  7. 2025Reward Forcing: Efficient Streaming Video Generation with Rewarded Distribution Matching Distillationcited by 86
  8. 2025Radial Attention: O(nlog n) Sparse Attention with Energy Decay for Long Video Generationcited by 69
  9. 2025ATLAS: Learning to Optimally Memorize the Context at Test Timecited by 66
  10. 2026Helios: Real Real-Time Long Video Generation Modelcited by 65
  11. 2025It's All Connected: A Journey Through Test-Time Memorization, Attentional Bias, Retention, and Online Optimizationcited by 64
  12. 2025End-to-End Test-Time Training for Long Contextcited by 62
38 more
  1. 2025Hunyuan-GameCraft-2: Instruction-following Interactive Game World Modelcited by 61
  2. 2025HoloCine: Holistic Generation of Cinematic Multi-Shot Long Video Narrativescited by 54
  3. 2025MemFlow: Flowing Adaptive Memory for Consistent and Efficient Long Video Narrativescited by 45
  4. 2026Lyra 2.0: Explorable Generative 3D Worldscited by 44
  5. 2025VISTA: A Test-Time Self-Improving Video Generation Agentcited by 31
  6. 2026Scal3R: Scalable Test-Time Training for Large-Scale 3D Reconstructioncited by 31
  7. 2025StreamDiffusionV2: A Streaming System for Dynamic and Interactive Video Generationcited by 30
  8. 2025CineTrans: Learning to Generate Videos with Cinematic Transitions via Masked Diffusion Modelscited by 29
  9. 2025Pack and Force Your Memory: Long-form and Consistent Video Generationcited by 29
  10. 2026VideoMemory: Toward Consistent Video Generation via Memory Integrationcited by 25
  11. 2025VeriThinker: Learning to Verify Makes Reasoning Model Efficientcited by 24
  12. 2026Relax Forcing: Relaxed KV-Memory for Consistent Long Video Generationcited by 24
  13. 2025Hierarchical Fine-grained Preference Optimization for Physically Plausible Video Generationcited by 21
  14. 2026In-Place Test-Time Trainingcited by 20
  15. 2025AniMaker: Multi-Agent Animated Storytelling with MCTS-Driven Clip Generationcited by 19
  16. 2025LoViC: Efficient Long Video Generation with Context Compressioncited by 19
  17. 2025MoGA: Mixture-of-Groups Attention for End-to-End Long Video Generationcited by 18
  18. 2025UltraGen: High-Resolution Video Generation with Hierarchical Attentioncited by 18
  19. 2025ViT3: Unlocking Test-Time Training in Visioncited by 18
  20. 2026VGG-T3: Offline Feed-Forward 3D Reconstruction at Scalecited by 18
  21. 2026MBench: A Comprehensive Benchmark on Memory Capability for Video World Modelscited by 18
  22. 2025FreeLong++: Training-Free Long Video Generation via Multi-band SpectralFusioncited by 17
  23. 2025Think Before You Diffuse: Infusing Physical Rules into Video Diffusioncited by 16
  24. 2025WorldWeaver: Generating Long-Horizon Video Worlds via Rich Perceptioncited by 16
  25. 2026Anchor Forcing: Anchor Memory and Tri-Region RoPE for Interactive Streaming Video Diffusioncited by 16
  26. 2025AnimaX: Animating the Inanimate in 3D with Joint Video-Pose Diffusion Modelscited by 15
  27. 2026AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memoriescited by 15
  28. 2026TokenGS: Decoupling 3D Gaussian Prediction from Pixels with Learnable Tokenscited by 15
  29. 20254D-LRM: Large Space-Time Reconstruction Model From and To Any View at Any Timecited by 14
  30. 2026Test-Time Training with KV Binding Is Secretly Linear Attentioncited by 13
  31. 2025DISCOVER: Automated Curricula for Sparse-Reward Reinforcement Learningcited by 12
  32. 2026Light Forcing: Accelerating Autoregressive Video Diffusion via Sparse Attentioncited by 12
  33. 2026MSVBench: Towards Human-Level Evaluation of Multi-Shot Video Generationcited by 12
  34. 2026Mode Seeking meets Mean Seeking for Fast Long Video Generationcited by 12
  35. 2026VQQA: An Agentic Approach for Video Evaluation and Quality Improvementcited by 12
  36. 2025FilmWeaver: Weaving Consistent Multi-Shot Videos with Cache-Guided Autoregressive Diffusioncited by 10
  37. 2025Local Mixtures of Experts: Essentially Free Test-Time Training via Model Mergingcited by 9
  38. 2025Revisiting Test-Time Scaling: A Survey and a Diversity-Aware Method for Efficient Reasoningcited by 9

50 most cited of 101 All 101 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2025MALT Diffusion: Memory-Augmented Latent Transformers for Any-Length Video Generation
  2. 2025Pushing the Boundaries of State Space Models for Image and Video Generation
  3. 2025Video-T1: Test-Time Scaling for Video Generation
  4. 2024xGen-VideoSyn-1: High-fidelity Text-to-Video Synthesis with Compressed Representations
  5. 2022MAGVIT: Masked Generative Video Transformer
  6. 2025ShotAdapter: Text-to-Multi-Shot Video Generation with Diffusion Models
  7. 2021VideoGPT: Video Generation using VQ-VAE and Transformers
  8. 2024DiTCtrl: Exploring Attention Control in Multi-Modal Diffusion Transformer for Tuning-Free Multi-Prompt Longer Video Generation