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 130 篇，第 2 / 3 页

51. NeurIPS 2026Git Context Controller: Manage the Context of Agents by Agentic Git
52. NeurIPS 2026Global Context Guidance for Diffusion Large Language Models
53. NeurIPS 2026Heterogeneous Parallelism for Multimodal Large Language Model Training
54. NeurIPS 2026Imitation Dominates Reinforcement: Direct In-Context RL Is Closer to ICL Than RL
55. NeurIPS 2026Improving Context-Shift Robustness of Convolutional Networks via Context-Regularized Cross-Entropy
56. NeurIPS 2026In-Context Benign Overfitting: A Feature-Selection Model in Linear Regression ICL
57. NeurIPS 2026In-context learning to predict critical transitions in dynamical systems
58. NeurIPS 2026Inadvertent Context Leakage in Language Models
59. NeurIPS 2026Incremental Learning in Transformers for In-Context Associative Recall
60. NeurIPS 2026Jet-Long: Efficient Long-Context Extension with Dynamic Bifocal RoPE
61. NeurIPS 2026Know Where You Stand: Memory-Source Choice in Long-Context Dialogue Agents
62. NeurIPS 2026KV Packet: Recomputation-Free Context-Independent KV Caching for LLMs
63. NeurIPS 2026Large Language Models Develop Belief State Geometry In-Context
64. NeurIPS 2026Learning to Follow In-Context Watermark Instructions via Self-Distillation
65. NeurIPS 2026Learning-to-Memorize: Dynamic Context Management for Long-Horizon Autoregressive Video Generation
66. NeurIPS 2026Lemon: Evidence-Risk-Aware Adaptive Organization for Long-Horizon LLM Agents
67. NeurIPS 2026LensVLM: Selective Context Expansion for Compressed Visual Representation of Text
68. NeurIPS 2026Linguistic Trajectory Encoding for Efficient Long-Horizon Spatial Memory in Embodied Agents
69. NeurIPS 2026Logarithmic Depth Suffices for In-Context Gradient Descent
70. NeurIPS 2026Logistic Bandits with $\tilde{O}(\sqrt{dT})$ Regret without Context Diversity Assumptions
71. NeurIPS 2026Long-Context Generation Is a Sampling Problem
72. NeurIPS 2026Long-Context Language Models Require Extreme Sparsity in Context Dimension
73. NeurIPS 2026Long-Horizon Agency Belongs in the Harness, Not the Context Window Only
74. NeurIPS 2026Long-Horizon Q-Learning: Accurate Value Learning via n-Step Inequalities
75. NeurIPS 2026Long-Term Risks of Risk-Based Allocation
76. NeurIPS 2026LongScape: Advancing Long-Horizon Embodied World Models with Context-Aware MoE
77. NeurIPS 2026Make Each Token Count: Towards Improving Long-Context Performance with KV Cache Eviction
78. NeurIPS 2026MedHorizon: Towards Long-context Medical Video Understanding in the Wild
79. NeurIPS 2026Motif-Mamba: network motif improved mamba for long-range sequence modeling
80. NeurIPS 2026MultiTalk: Scaling Full-Duplex Speech Models to Long, Multi-Party, Bilingual Conversation
81. NeurIPS 2026Neuromodulated Constrained Autoencoders for Context-Dependent Manifold Learning
82. NeurIPS 2026NINJA: A Navigator–Inspector Joint Architecture for Context-Efficient Issue Localization
83. NeurIPS 2026NOCE-Net: Representation Learning for Battery Operational Context via Nested Sequence Modelling
84. NeurIPS 2026Not All Tokens Should Be Treated Equally: Context Credits Reassignment
85. NeurIPS 2026OmniMemBench: Towards Scalable Evaluation of Long-Term Omni-Modal Agent Memory
86. NeurIPS 2026On the Importance of Gating: Memorization vs. In-Context Learning in State Space Models
87. NeurIPS 2026Optimal Contextual Pricing under Agnostic Non-Lipschitz Demand
88. NeurIPS 2026PEEK: Context Map as an Orientation Cache for Long-Context LLM Agents
89. NeurIPS 2026PermaVid: Consistent Video Generation Across Edits via Disentangled Context Memory
90. NeurIPS 2026Personalize-then-Store: Benchmarking and Learning Personalized Memory for Long-Horizon Agents
91. NeurIPS 2026Point4D: Long-range 4D Motion Reconstruction
92. NeurIPS 2026Preventing Error Cascades in Long-Horizon Multimodal Agents with Edge-Reliability Graph Memory
93. NeurIPS 2026PRISM: Primitive Routing via In-context Skill Mixing for Lifelong VLA
94. NeurIPS 2026Pure Exploration Beyond Reward Feedback: The Role of Post-Action Context
95. NeurIPS 2026Q-Focus: Let the Question Guide What to See in Long Videos
96. NeurIPS 2026Recursive Language Models
97. NeurIPS 2026Resilient Latent Readouts for Long-Context Question Answering
98. NeurIPS 2026RoadmapBench: Evaluating Long-Horizon Agentic Software Development Across Version Upgrades
99. NeurIPS 2026ROLLVERIFY: BRIDGING EFFICIENCY AND ACCURACY IN LONG-TAIL ROLLOUT REINFORCEMENT LEARNING
100. NeurIPS 2026SaaSBench: Exploring the Boundaries of Coding Agents in Long-Horizon Enterprise SaaS Engineering

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。