长上下文论文趋势 2023–2026

主要结论

• 长上下文论文从 2023 年的 66 篇增长到 2026 年的 341 篇，三年间扩大到约 5.2 倍。
• NeurIPS 是发长上下文论文最多的会议，2023–2026 年共 298 篇，占四个会议总量的 36%。
• 交叉方向里涨得最快的是AI 智能体：同时属于两个主题的论文从 2023 年的 0 篇增至 2026 年的 45 篇。
• 在「长上下文与记忆」的 5 个主题里，它的 2026 年论文数排第 1；同组增长最快的检索记忆是 6.6 倍，长上下文是 5.2 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 2026Accelerating Long-Context LLM Prefill via Layer-wise Progressive Token Pruning in Local Deployment
• ICML 2026AdaSplash-2: Faster Differentiable Sparse Attention
• ICLR 2026An evolutionary perspective on modes of learning in Transformers
• CVPR 2026AR2-4FV Anchored Referring and Re-identification for Long-Term Grounding in Fixed-View
• NeurIPS 2025A Controllable Examination for Long-Context Language Models
• ICML 2025CateKV: On Sequential Consistency for Long-Context LLM Inference Acceleration
• ICLR 2025$\text{D}_{2}\text{O}$: Dynamic Discriminative Operations for Efficient Long-Context Inference of Large Language Models
• CVPR 2025BIMBA: Selective-Scan Compression for Long-Range Video Question Answering
• NeurIPS 2024[Re] CUDA: Curriculum of Data Augmentation for Long‐tailed Recognition
• ICML 2024Benchmarking and Building Long-Context Retrieval Models with LoCo and M2-BERT
• ICLR 2024CLEX: Continuous Length Extrapolation for Large Language Models
• CVPR 2024GigaTraj: Predicting Long-term Trajectories of Hundreds of Pedestrians in Gigapixel Complex Scenes
• NeurIPS 2023A Single 2D Pose with Context is Worth Hundreds for 3D Human Pose Estimation
• ICML 2023CAB: Comprehensive Attention Benchmarking on Long Sequence Modeling
• ICLR 2023$k$NN Prompting: Beyond-Context Learning with Calibration-Free Nearest Neighbor Inference
• CVPR 2023AutoAD: Movie Description in Context
• NeurIPS 2026AdaCal: Adaptive Calibration for Robust Sparse Attention in Long-Context LLMs
• ICML 2026Artificial Hippocampus Networks for Efficient Long-Context Modeling
• ICLR 2026AutoSP: Unlocking Long-Context LLM Training Via Compiler-Based Sequence Parallelism
• CVPR 2026AstraNav-Memory Contexts Compression for Long Memory

全部长上下文论文

共 53 篇，第 1 / 2 页

1. ICML 2025$\infty$-Video: A Training-Free Approach to Long Video Understanding via Continuous-Time Memory Consolidation
2. ICML 2025AdaSplash: Adaptive Sparse Flash Attention
3. ICML 2025Advancing Personalized Learning with Neural Collapse for Long-Tail Challenge
4. ICML 2025Behavioral Exploration: Learning to Explore via In-Context Adaptation
5. ICML 2025Can Transformers Learn Full Bayesian Inference in Context?
6. ICML 2025CateKV: On Sequential Consistency for Long-Context LLM Inference Acceleration
7. ICML 2025CoMemo: LVLMs Need Image Context with Image Memory
8. ICML 2025Context Matters: Query-aware Dynamic Long Sequence Modeling of Gigapixel Images
9. ICML 2025Core Context Aware Transformers for Long Context Language Modeling
10. ICML 2025Curse of High Dimensionality Issue in Transformer for Long Context Modeling
11. ICML 2025Customizing the Inductive Biases of Softmax Attention using Structured Matrices
12. ICML 2025Decision Mixer: Integrating Long-term and Local Dependencies via Dynamic Token Selection for Decision-Making
13. ICML 2025Does learning the right latent variables necessarily improve in-context learning?
14. ICML 2025Efficient Length-Generalizable Attention via Causal Retrieval for Long-Context Language Modeling
15. ICML 2025Efficient Long Context Fine-tuning with Chunk Flow
16. ICML 2025Efficient Multi-modal Long Context Learning for Training-free Adaptation
17. ICML 2025EPIC: Efficient Position-Independent Caching for Serving Large Language Models
18. ICML 2025Everything Everywhere All at Once: LLMs can In-Context Learn Multiple Tasks in Superposition
19. ICML 2025Geometric Hyena Networks for Large-scale Equivariant Learning
20. ICML 2025In-Context Deep Learning via Transformer Models
21. ICML 2025In-Context Reinforcement Learning From Suboptimal Historical Data
22. ICML 2025Iterative Vectors: In-Context Gradient Steering without Backpropagation
23. ICML 2025LaCache: Ladder-Shaped KV Caching for Efficient Long-Context Modeling of Large Language Models
24. ICML 2025Learning In-context $n$-grams with Transformers: Sub-$n$-grams Are Near-Stationary Points
25. ICML 2025LOGO --- Long cOntext aliGnment via efficient preference Optimization
26. ICML 2025Long-Short Alignment for Effective Long-Context Modeling in LLMs
27. ICML 2025LongRoPE2: Near-Lossless LLM Context Window Scaling
28. ICML 2025LV-XAttn: Distributed Cross-Attention for Long Visual Inputs in Multimodal Large Language Models
29. ICML 2025M+: Extending MemoryLLM with Scalable Long-Term Memory
30. ICML 2025MMInference: Accelerating Pre-filling for Long-Context Visual Language Models via Modality-Aware Permutation Sparse Attention
31. ICML 2025NExtLong: Toward Effective Long-Context Training without Long Documents
32. ICML 2025NoLiMa: Long-Context Evaluation Beyond Literal Matching
33. ICML 2025On Measuring Long-Range Interactions in Graph Neural Networks
34. ICML 2025On-the-Fly Adaptive Distillation of Transformer to Dual-State Linear Attention for Long-Context LLM Serving
35. ICML 2025ParallelComp: Parallel Long-Context Compressor for Length Extrapolation
36. ICML 2025PENCIL: Long Thoughts with Short Memory
37. ICML 2025Provable In-Context Vector Arithmetic via Retrieving Task Concepts
38. ICML 2025QuantSpec: Self-Speculative Decoding with Hierarchical Quantized KV Cache
39. ICML 2025RAPID: Long-Context Inference with Retrieval-Augmented Speculative Decoding
40. ICML 2025Rethinking Addressing in Language Models via Contextualized Equivariant Positional Encoding
41. ICML 2025Rethinking Benign Overfitting in Two-Layer Neural Networks
42. ICML 2025RocketKV: Accelerating Long-Context LLM Inference via Two-Stage KV Cache Compression
43. ICML 2025ShadowKV: KV Cache in Shadows for High-Throughput Long-Context LLM Inference
44. ICML 2025Star Attention: Efficient LLM Inference over Long Sequences
45. ICML 2025Strategy Coopetition Explains the Emergence and Transience of In-Context Learning
46. ICML 2025TabICL: A Tabular Foundation Model for In-Context Learning on Large Data
47. ICML 2025Test-Time Training Provably Improves Transformers as In-context Learners
48. ICML 2025TimeBridge: Non-Stationarity Matters for Long-term Time Series Forecasting
49. ICML 2025Training Dynamics of In-Context Learning in Linear Attention
50. ICML 2025TypyBench: Evaluating LLM Type Inference for Untyped Python Repositories

常见问题

2023–2026 年四大会议一共有多少篇长上下文论文？

共 838 篇，按年份分别是 2023 年 66 篇、2024 年 176 篇、2025 年 255 篇、2026 年 341 篇。

NeurIPS 2026 有多少篇长上下文论文？

NeurIPS 2026 有 130 篇长上下文论文，上一届是 98 篇。

ICLR 2026 有多少篇长上下文论文？

ICLR 2026 有 82 篇长上下文论文，上一届是 87 篇。

ICML 2026 有多少篇长上下文论文？

ICML 2026 有 98 篇长上下文论文，上一届是 53 篇。

CVPR 2026 有多少篇长上下文论文？

CVPR 2026 有 31 篇长上下文论文，上一届是 17 篇。

哪个会议的长上下文论文最多？

NeurIPS，2023–2026 年共 298 篇。