mlsanityData to 7 Oct 2026

MemFlow: Flowing Adaptive Memory for Consistent and Efficient Long Video Narratives

arXiv 2512.14699 ↗First posted 16 Dec 2025cs.CVcited by 40

Not on the map. Why?

Abstract (arXiv)

The core challenge for streaming video generation is maintaining the content consistency in long context, which poses high requirement for the memory design. Most existing solutions maintain the memory by compressing historical frames with predefined strategies. However, different to-generate video chunks should refer to different historical cues, which is hard to satisfy with fixed strategies. In this work, we propose MemFlow to address this problem. Specifically, before generating the coming chunk, we dynamically update the memory bank by retrieving the most relevant historical frames with the text prompt of this chunk. This design enables narrative coherence even if new event happens or scenario switches in future frames. In addition, during generation, we only activate the most relevant tokens in the memory bank for each query in the attention layers, which effectively guarantees the generation efficiency. In this way, MemFlow achieves outstanding long-context consistency with negligible computation burden (7.9% speed reduction compared with the memory-free baseline) and keeps the compatibility with any streaming video generation model with KV cache.

Key facts. Cited by 40 papers in the map (data through 7 Oct 2026). Builds on 41 of its 46 references that are in the map. First posted on arXiv 16 Dec 2025.

Builds on

41 of 46 references.

References · 41

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  2. 2022Scalable Diffusion Models with Transformerscited by 5,931
  3. 2025Wan: Open and Advanced Large-Scale Video Generative Modelscited by 2,536
  4. 2024CogVideoX: Text-to-Video Diffusion Models with An Expert Transformercited by 2,198
  5. 2024Qwen2 Technical Reportcited by 2,076
  6. 2023VBench: Comprehensive Benchmark Suite for Video Generative Modelscited by 1,755
  7. 2024HunyuanVideo: A Systematic Framework For Large Video Generative Modelscited by 1,597
  8. 2023One-step Diffusion with Distribution Matching Distillationcited by 978
  9. 2024Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusioncited by 634
  10. 2024An Image is Worth 1/2 Tokens After Layer 2: Plug-and-Play Inference Acceleration for Large Vision-Language Modelscited by 629
  11. 2025Self Forcing: Bridging the Train-Test Gap in Autoregressive Video Diffusioncited by 569
  12. 2024LTX-Video: Realtime Video Latent Diffusioncited by 549
29 more
  1. 2024From Slow Bidirectional to Fast Autoregressive Video Diffusion Modelscited by 401
  2. 2024Pyramidal Flow Matching for Efficient Video Generative Modelingcited by 354
  3. 2025MAGI-1: Autoregressive Video Generation at Scalecited by 262
  4. 2024StreamingT2V: Consistent, Dynamic, and Extendable Long Video Generation from Textcited by 242
  5. 2024StoryDiffusion: Consistent Self-Attention for Long-Range Image and Video Generationcited by 240
  6. 2024MA-LMM: Memory-Augmented Large Multimodal Model for Long-Term Video Understandingcited by 223
  7. 2025SkyReels-V2: Infinite-length Film Generative Modelcited by 218
  8. 2024Autoregressive Video Generation without Vector Quantizationcited by 180
  9. 2025Context as Memory: Scene-Consistent Interactive Long Video Generation with Memory Retrievalcited by 157
  10. 2024In-Context LoRA for Diffusion Transformerscited by 155
  11. 2025WorldMem: Long-term Consistent World Simulation with Memorycited by 152
  12. 2025Test-Time Training Done Rightcited by 121
  13. 2025Long-Context Autoregressive Video Modeling with Next-Frame Predictioncited by 119
  14. 2025VMem: Consistent Interactive Video Scene Generation with Surfel-Indexed View Memorycited by 102
  15. 2025One-Minute Video Generation with Test-Time Trainingcited by 100
  16. 2025Mixture of Contexts for Long Video Generationcited by 90
  17. 2024MovieDreamer: Hierarchical Generation for Coherent Long Visual Sequencecited by 85
  18. 2024Vlogger: Make Your Dream A Vlogcited by 70
  19. 2024VideoStudio: Generating Consistent-Content and Multi-Scene Videoscited by 62
  20. 2025StreamMem: Query-Agnostic KV Cache Memory for Streaming Video Understandingcited by 53
  21. 2025Captain Cinema: Towards Short Movie Generationcited by 48
  22. 2025AdaReTaKe: Adaptive Redundancy Reduction to Perceive Longer for Video-language Understandingcited by 45
  23. 2025VRAG: Learning World Models for Interactive Video Generationcited by 37
  24. 2024SlowFast-VGen: Slow-Fast Learning for Action-Driven Long Video Generationcited by 28
  25. 2025FilMaster: Bridging Cinematic Principles and Generative AI for Automated Film Generationcited by 28
  26. 2025StarGen: A Spatiotemporal Autoregression Framework with Video Diffusion Model for Scalable and Controllable Scene Generationcited by 26
  27. 2025LoViC: Efficient Long Video Generation with Context Compressioncited by 19
  28. 2025Macro-from-Micro Planning for High-Quality and Parallelized Autoregressive Long Video Generationcited by 19
  29. 2024Beyond Training: Dynamic Token Merging for Zero-Shot Video Understandingcited by 13

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 40.

40 papers.

Sort:

By area:

  1. 2026Relax Forcing: Relaxed KV-Memory for Consistent Long Video Generationcited by 20
  2. 2026MBench: A Comprehensive Benchmark on Memory Capability for Video World Modelscited by 15
  3. 2026Anchor Forcing: Anchor Memory and Tri-Region RoPE for Interactive Streaming Video Diffusioncited by 14
  4. 2026WildWorld: A Large-Scale Dataset for Dynamic World Modeling with Actions and Explicit State toward Generative ARPGcited by 13
  5. 2026Pathwise Test-Time Correction for Autoregressive Long Video Generationcited by 12
  6. 2026Mode Seeking meets Mean Seeking for Fast Long Video Generationcited by 12
  7. 2026KVPO: ODE-Native GRPO for Autoregressive Video Alignment via KV Semantic Explorationcited by 9
  8. 2026CausalCine: Real-Time Autoregressive Generation for Multi-Shot Video Narrativescited by 8
  9. 2026Long-Horizon Streaming Video Generation via Hybrid Attention with Decoupled Distillationcited by 7
  10. 2026Efficient Video Diffusion Models: Advancements and Challengescited by 6
  11. 2026RAVEN: Real-time Autoregressive Video Extrapolation with Consistency-model GRPOcited by 6
  12. 2026Advancing Narrative Long Video Generation via Training-Free Identity-Aware Memorycited by 6
28 more
  1. 2026Echo-Infinity: Learning Evolving Memory for Real-Time Infinite Video Generationcited by 6
  2. 2026Video Generation Models as World Models: Efficient Paradigms, Architectures and Algorithmscited by 5
  3. 2026Head Forcing: Long Autoregressive Video Generation via Head Heterogeneitycited by 5
  4. 2026DySink: Dynamic Frame Sinks for Autoregressive Long Video Generationcited by 5
  5. 2026SlotMemory: Object-Centric KV Memory for Streaming Long-Video Generationcited by 5
  6. 2026CineDance: Towards Next-Generation Multi-Shot Long-Form Cinematic Audio-Video Generationcited by 5
  7. 2026FadeMem: Distance-Aware Memory Consolidation for Autoregressive Video Diffusioncited by 5
  8. 2026DIM-WAM: World-Action Modeling with Diverse Historical Event Memorycited by 5
  9. 2026Adapting VACE for Real-Time Autoregressive Video Diffusioncited by 4
  10. 2026Echo-Memory: A Controlled Study of Memory in Action World Modelscited by 4
  11. 2026TempAct: Advancing Temporal Plausibility in Autoregressive Video Generation via Planner-Executor RLcited by 4
  12. 2026StreamAV-Bench: A Comprehensive Benchmark for Streaming Audio-Video Generationcited by 4
  13. 2026CoDance: An Unbind-Rebind Paradigm for Robust Multi-Subject Animationcited by 3
  14. 2026Research on World Models Is Not Merely Injecting World Knowledge into Specific Taskscited by 3
  15. 2026Goodbye Drift: Anchored Tree Sampling for Long-Horizon Video-to-Video Generationcited by 2
  16. 2026StreamEdit: Training-Free Video Editing via Few-Step Streaming Video Generationcited by 2
  17. 2026OmniMem: Scalable and Adaptive Memory Retrieval for Long Video Generationcited by 2
  18. 2026Surprise Forcing: What to Remember, When to Skip in Long Video Generationcited by 2
  19. 2026StreamHOI: Interaction-aware Temporal Memory Adaptation for Streaming HOI Video Generationcited by 2
  20. 2026Visko Orbis 1.0: A Live Model for Real-Time Interactive Long Video Generationcited by 2
  21. 2026RECAP-Forcing: Retaining Content Appearances for Long Video Generationcited by 2
  22. 2026LayerRecall: A State-Conditioned Memory Router for Long-Horizon Consistency in Video Generationcited by 2
  23. 2026Delta Forcing: Trust Region Steering for Interactive Autoregressive Video Generationcited by 1
  24. 2026SlotMem: Character-Addressable Internal Memory for Narrative Long Video Generationcited by 1
  25. 2026SEAM: Shot Entity-Attribute Memory for Consistent Short-Drama Generation at Scalecited by 1
  26. 2026Self-Aligned Forcing: Streaming Video Diffusion with Differentiable Noisy Historycited by 0
  27. 2026No Corners Cut: State-Grounded Transitions for Mid-Stream Prompt Switches in Video Generationcited by 0
  28. 2026FrameMorrow: Future-guided Frame Selection with Prospective Tokens for Long-Horizon Video Generationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Memorize-and-Generate: Towards Long-Term Consistency in Real-Time Video Generation
  2. 2025VideoSSM: Autoregressive Long Video Generation with Hybrid State-Space Memory
  3. 2026MemCam: Memory-Augmented Camera Control for Consistent Video Generation
  4. 2026Towards Error-Free Long Video Generation
  5. 2026Memory Forcing: Attendable Mid-Horizon History for Streaming Video Generation
  6. 2026SWIFT: Prompt-Adaptive Memory for Efficient Interactive Long Video Generation
  7. 2026StreamFlow: Dynamic Memory Flows for Streaming Video Understanding
  8. 2026FluxMem: Adaptive Hierarchical Memory for Streaming Video Understanding