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 82 篇，第 1 / 2 页

1. ICLR 2026Adversarially Pretrained Transformers may be Universally Robust In-Context Learners
2. ICLR 2026AgentFold: Long-Horizon Web Agents with Proactive Context Folding
3. ICLR 2026Agentic Context Engineering: Learning Comprehensive Contexts for Self-Improving Language Models
4. ICLR 2026AMemGym: Interactive Memory Benchmarking for Assistants in Long-Horizon Conversations
5. ICLR 2026An evolutionary perspective on modes of learning in Transformers
6. ICLR 2026Autoencoding-Free Context Compression for LLMs via Contextual Semantic Anchors
7. ICLR 2026AutoSP: Unlocking Long-Context LLM Training Via Compiler-Based Sequence Parallelism
8. ICLR 2026Bayesian Attention Mechanism: A Probabilistic Framework for Positional Encoding and Context Length Extrapolation
9. ICLR 2026Beyond Length: Quantifying Long-Range Information for Long-Context LLM Pretraining Data
10. ICLR 2026Beyond RAG vs. Long-Context: Learning Distraction-Aware Retrieval for Efficient Knowledge Grounding
11. ICLR 2026Beyond Real: Imaginary Extension of Rotary Position Embeddings for Long-Context LLMs
12. ICLR 2026Cache What Lasts: Token Retention for Memory-Bounded KV Cache in LLMs
13. ICLR 2026Cartridges: Lightweight and general-purpose long context representations via self-study
14. ICLR 2026CDBridge: A Cross-omics Post-training Bridge Strategy for Context-aware Biological Modeling
15. ICLR 2026Channel-Aware Mixed-Precision Quantization for Efficient Long-Context Inference
16. ICLR 2026COMI: Coarse-to-fine Context Compression via Marginal Information Gain
17. ICLR 2026Context and Diversity Matter: The Emergence of In-Context Learning in World Models
18. ICLR 2026ContextNav: Towards Agentic Multimodal In-Context Learning
19. ICLR 2026Continuum Transformers Perform In-Context Learning by Operator Gradient Descent
20. ICLR 2026Critical attention scaling in long-context transformers
21. ICLR 2026Draft-based Approximate Inference for LLMs
22. ICLR 2026DSSA: Dense-Sparse Switchable Attention for Seamless Short-to-Long Adaptation
23. ICLR 2026ECHO: Toward Contextual Seq2Seq Paradigms in Large EEG Models
24. ICLR 2026EntropyLong: Effective Long-Context Training via Predictive Uncertainty
25. ICLR 2026Extending the Context of Pretrained LLMs by Dropping Their Positional Embedding
26. ICLR 2026Frayed RoPE and Long Inputs: A Geometric Perspective
27. ICLR 2026FreqKV: Key-Value Compression in Frequency Domain for Context Window Extension
28. ICLR 2026From Collapse to Control: Understanding and Extending Context Length in Emerging Hybrid Models via Universal Position Interpolation
29. ICLR 2026From Markov to Laplace: How Mamba In-Context Learns Markov Chains
30. ICLR 2026FSA: An Alternative Efficient Implementation of Native Sparse Attention Kernel
31. ICLR 2026Generative Diffusion Prior Distillation for Long-Context Knowledge Transfer
32. ICLR 2026Improving Long-Range Interactions in Graph Neural Simulators via Hamiltonian Dynamics
33. ICLR 2026In-Context Algebra
34. ICLR 2026In-Context Algorithm Emulation in Fixed-Weight Transformers
35. ICLR 2026Inferring brain plasticity rule under long-term stimulation with structured recurrent dynamics
36. ICLR 2026InftyThink: Breaking the Length Limits of Long-Context Reasoning in Large Language Models
37. ICLR 2026Intrinsic Entropy of Context Length Scaling in LLMs
38. ICLR 2026Is In-Context Learning Learning?
39. ICLR 2026IterResearch: Rethinking Long-Horizon Agents via Markovian State Reconstruction
40. ICLR 2026Learning to Adapt: In-Context Learning Beyond Stationarity
41. ICLR 2026Let's (not) just put things in Context: Test-time Training for Long-context LLMs
42. ICLR 2026Long-Context Attention Benchmark: From Kernel Efficiency to Distributed Context Parallelism
43. ICLR 2026Long-Context Generalization with Sparse Attention
44. ICLR 2026Long-range Modeling and Processing of Multimodal Event Sequences
45. ICLR 2026LongRLVR: Long-Context Reinforcement Learning Requires Verifiable Context Rewards
46. ICLR 2026LoongRL: Reinforcement Learning for Advanced Reasoning over Long Contexts
47. ICLR 2026LycheeDecode: Accelerating Long-Context LLM Inference via Hybrid-Head Sparse Decoding
48. ICLR 2026MemAgent: Reshaping Long-Context LLM with Multi-Conv RL-based Memory Agent
49. ICLR 2026Mini-cluster Guided Long-tailed Deep Clustering
50. ICLR 2026MMReD: a Cross-Modal Benchmark for Dense Context Reasoning

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。