叙事生成论文趋势 2023–2026

主要结论

• 叙事生成论文从 2023 年的 4 篇增长到 2026 年的 61 篇，三年间扩大到约 15 倍。
• NeurIPS 是发叙事生成论文最多的会议，2023–2026 年共 33 篇，占四个会议总量的 33%。
• 交叉方向里涨得最快的是文生视频：同时属于两个主题的论文从 2023 年的 0 篇增至 2026 年的 22 篇。
• 在「多模态生成」的 5 个主题里，它的 2026 年论文数排第 4；同组增长最快的视听生成是 3.7 倍，叙事生成是 15 倍。

代表论文

按年份和会议均匀挑选，方便快速看到每一届的样貌

• NeurIPS 2026AdaPCLA: Curriculum Prior Internalization For Long-Tailed Longitudinal EHR Generation
• ICML 2026Controlled SDEs for Long-Horizon Motion Generation under Latent Decision Uncertainty
• ICLR 2026ExpertLongBench: Benchmarking Language Models on Expert-Level Long-Form Generation Tasks with Structured Checklists
• CVPR 2026Endless World Real-Time 3D-Aware Long Video Generation
• NeurIPS 2025Deep Compositional Phase Diffusion for Long Motion Sequence Generation
• ICML 2025TokenSwift: Lossless Acceleration of Ultra Long Sequence Generation
• ICLR 2025LongGenBench: Benchmarking Long-Form Generation in Long Context LLMs
• CVPR 2025FreeCloth: Free-form Generation Enhances Challenging Clothed Human Modeling
• NeurIPS 2024LVD-2M: A Long-take Video Dataset with Temporally Dense Captions
• ICML 2024Viewing Transformers Through the Lens of Long Convolutions Layers
• ICLR 2024Achieving Human Parity in Content-Grounded Datasets Generation
• CVPR 2024MM-Narrator: Narrating Long-form Videos with Multimodal In-Context Learning
• NeurIPS 2023Unlimiformer: Long-Range Transformers with Unlimited Length Input
• ICLR 2023Long-Tailed Partial Label Learning via Dynamic Rebalancing
• CVPR 2023Relational Space-Time Query in Long-Form Videos
• NeurIPS 2026Memorize When Needed: Decoupled Memory Control for Spatially Consistent Long-Horizon Video Generation
• ICML 2026Embedding Trust: Semantic Isotropy Predicts Nonfactuality in Long-Form Text Generation
• ICLR 2026LongLive: Real-time Interactive Long Video Generation
• CVPR 2026HoloCine Holistic Generation of Cinematic Multi-Shot Long Video Narratives
• NeurIPS 2025Keep It on a Leash: Controllable Pseudo-label Generation Towards Realistic Long-Tailed Semi-Supervised Learning

全部叙事生成论文

共 100 篇，第 1 / 2 页

1. CVPR 2026Adaptive Auxiliary Prompt Blending for Target-Faithful Diffusion Generation
2. CVPR 2026Decision Boundary-aware Generation for Long-tailed Learning
3. CVPR 2026Endless World Real-Time 3D-Aware Long Video Generation
4. CVPR 2026Free-Lunch Long Video Generation via Layer-Adaptive O.O.D Correction
5. CVPR 2026HoloCine Holistic Generation of Cinematic Multi-Shot Long Video Narratives
6. CVPR 2026Infinity-RoPE Action-Controllable Infinite Video Generation Emerges From Autoregressive Self-Rollout
7. CVPR 2026Learning Long-term Motion Embeddings for Efficient Kinematics Generation
8. CVPR 2026LoL Longer than Longer Scaling Video Generation to Hour
9. CVPR 2026Long-SCOPE Fully Sparse Long-Range Cooperative 3D Perception
10. CVPR 2026MANSION Multi-floor lANguage-to-3D Scene generatIOn for loNg-horizon tasks
11. CVPR 2026OneStory Coherent Multi-Shot Video Generation with Adaptive Memory
12. CVPR 2026PerpetualWonder Long-horizon Action-conditioned 4D Scene Generation
13. CVPR 2026Reframing Long-Tailed Learning via Loss Landscape Geometry
14. CVPR 2026ShotDirector Directorially Controllable Multi-Shot Video Generation with Cinematographic Transitions
15. CVPR 2026TempoMaster Efficient Long Video Generation via Next-Frame-Rate Prediction
16. CVPR 2026VinQA Visual Elements Interleaved Long-form Answer Generation for Real-World Multimodal
17. ICLR 2026ExpertLongBench: Benchmarking Language Models on Expert-Level Long-Form Generation Tasks with Structured Checklists
18. ICLR 2026Leveraging Pretrained Knowledge at Inference Time: LoRA-Gated Contrastive Decoding for Multilingual Factual Language Generation in Adapted LLMs
19. ICLR 2026Long-Text-to-Image Generation via Compositional Prompt Decomposition
20. ICLR 2026LongLive: Real-time Interactive Long Video Generation
21. ICLR 2026LongWriter-Zero: Mastering Ultra-Long Text Generation via Reinforcement Learning
22. ICLR 2026MAGE: Multi-scale Autoregressive Generation for Offline Reinforcement Learning
23. ICLR 2026Micro-Macro Retrieval: Reducing Long-Form Hallucination in Large Language Models
24. ICLR 2026MoGA: Mixture-of-Groups Attention for End-to-End Long Video Generation
25. ICLR 2026NarrLV: Towards a Comprehensive Narrative-Centric Evaluation for Long Video Generation
26. ICLR 2026PairFlow: Closed-Form Source-Target Coupling for Few-Step Generation in Discrete Flow Models
27. ICLR 2026ReactDance: Hierarchical Representation for High-Fidelity and Coherent Long-Form Reactive Dance Generation
28. ICLR 2026Reinforcement Learning from Dynamic Critic Feedback for Free-Form Generations
29. ICLR 2026Rolling Forcing: Autoregressive Long Video Diffusion in Real Time
30. ICLR 2026SketchingReality: From Freehand Scene Sketches to Photorealistic Images
31. ICLR 2026SteinsGate: Adding Causality to Diffusions for Long Video Generation via Path Integral
32. ICLR 2026Think-While-Generating: On-the-Fly Reasoning for Personalized Long-Form Generation
33. ICLR 2026YuE: Scaling Open Foundation Models for Long-Form Music Generation
34. ICML 2026Building Reliable Long-Form Generation via Hallucination Rejection Sampling
35. ICML 2026Controlled SDEs for Long-Horizon Motion Generation under Latent Decision Uncertainty
36. ICML 2026Embedding Trust: Semantic Isotropy Predicts Nonfactuality in Long-Form Text Generation
37. ICML 2026Enhancing Train-Free Infinite-Frame Generation for Consistent Long Videos
38. ICML 2026Evaluating LLM Uncertainty in Long-Form Generation Using Deterministic Ground Truth
39. ICML 2026MSP: Probabilistically Consistent Multi-Scale Action Generation
40. ICML 2026NL2Repo-Bench: Towards Long-Horizon Repository Generation Evaluation of Coding Agents
41. ICML 2026On Stable Long-Form Generation: Benchmarking and Mitigating Length Volatility
42. ICML 2026TextAtlas5M: A Large-Scale Dataset for Long Text Image Generation
43. NeurIPS 2026AdaPCLA: Curriculum Prior Internalization For Long-Tailed Longitudinal EHR Generation
44. NeurIPS 2026AgentOdyssey: Open-Ended Long-Horizon Text Game Generation for Test-Time Continual Learning Agents
45. NeurIPS 2026CausalCine: Real-Time Autoregressive Generation for Multi-Shot Video Narratives
46. NeurIPS 2026Controlling Transient Amplification Improves Long-horizon Rollouts
47. NeurIPS 2026Diffusion Thinking for Fast Long-Form Spatial Reasoning in Vision--Language Models
48. NeurIPS 2026GeoMemory: Geometry-Indexed Memory for Long-Horizon Interactive Video Generation
49. NeurIPS 2026Learning-to-Memorize: Dynamic Context Management for Long-Horizon Autoregressive Video Generation
50. NeurIPS 2026LogicTree-RAG: Logic Tree-guided Retrieval-Augmented Generation for Long-form Patent Drafting

常见问题

2023–2026 年四大会议一共有多少篇叙事生成论文？

共 100 篇，按年份分别是 2023 年 4 篇、2024 年 11 篇、2025 年 24 篇、2026 年 61 篇。

NeurIPS 2026 有多少篇叙事生成论文？

NeurIPS 2026 有 19 篇叙事生成论文，上一届是 9 篇。

ICLR 2026 有多少篇叙事生成论文？

ICLR 2026 有 17 篇叙事生成论文，上一届是 5 篇。

ICML 2026 有多少篇叙事生成论文？

ICML 2026 有 9 篇叙事生成论文，上一届是 3 篇。

CVPR 2026 有多少篇叙事生成论文？

CVPR 2026 有 16 篇叙事生成论文，上一届是 7 篇。

哪个会议的叙事生成论文最多？

NeurIPS，2023–2026 年共 33 篇。