上下文压缩论文趋势 2023–2026

主要结论

• 上下文压缩论文从 2023 年的 39 篇增长到 2026 年的 136 篇，三年间扩大到约 3.5 倍。
• NeurIPS 是发上下文压缩论文最多的会议，2023–2026 年共 123 篇，占四个会议总量的 34%。
• 交叉方向里涨得最快的是深度学习 · large language models：同时属于两个主题的论文从 2023 年的 1 篇增至 2026 年的 13 篇。
• 在「长上下文与记忆」的 5 个主题里，它的 2026 年论文数排第 4；同组增长最快的检索记忆是 6.6 倍，上下文压缩是 3.5 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 2026A Free Lunch in LLM Compression: Revisiting Retraining after Pruning
• ICML 2026A Unified Sparse Attention via Multi-Granularity Compression
• ICLR 2026An evolutionary perspective on modes of learning in Transformers
• CVPR 2026AstraNav-Memory Contexts Compression for Long Memory
• NeurIPS 2025AdmTree: Compressing Lengthy Context with Adaptive Semantic Trees
• ICML 2025Does learning the right latent variables necessarily improve in-context learning?
• ICLR 2025Approaching Rate-Distortion Limits in Neural Compression with Lattice Transform Coding
• CVPR 2025ECVC: Exploiting Non-Local Correlations in Multiple Frames for Contextual Video Compression
• NeurIPS 2024StreamingDialogue: Prolonged Dialogue Learning via Long Context Compression with Minimal Losses
• ICML 2024Agnostic Sample Compression Schemes for Regression
• ICLR 2024Compressed Context Memory for Online Language Model Interaction
• CVPR 2024Diversity-aware Channel Pruning for StyleGAN Compression
• NeurIPS 2023Do Not Marginalize Mechanisms, Rather Consolidate!
• ICML 2023Private Federated Learning with Autotuned Compression
• ICLR 2023$k$NN Prompting: Beyond-Context Learning with Calibration-Free Nearest Neighbor Inference
• CVPR 2023AutoAD: Movie Description in Context
• NeurIPS 2026A Self-Evolving Framework for Efficient Terminal Agents via Observational Context Compression
• ICML 2026Context-Driven Incremental Compression for Multi-Turn Dialogue Generation
• ICLR 2026Autoencoding-Free Context Compression for LLMs via Contextual Semantic Anchors
• CVPR 2026Featurising Pixels from Dynamic 3D Scenes with Linear In-Context Learners

全部上下文压缩论文

共 39 篇，第 1 / 1 页

1. ICML 2026A Unified Sparse Attention via Multi-Granularity Compression
2. ICML 2026ACON: Optimizing Context Compression for Long-horizon LLM Agents
3. ICML 2026Advancing SVD-based LLM Compression via Layer-Wise Error Model Search
4. ICML 2026Auto-regressive In-context Demonstration Selection
5. ICML 2026CGSVD: Cascaded Granular Singular Value Decomposition for Large Language Model Compression
6. ICML 2026CLAA: Cross-Layer Attention Aggregation for Accelerating LLM Prefill
7. ICML 2026ConceptMoE: Adaptive Token-to-Concept Compression for Implicit Compute Allocation
8. ICML 2026Context-Driven Incremental Compression for Multi-Turn Dialogue Generation
9. ICML 2026Context-free Recognition with Transformers
10. ICML 2026Efficient Learned Image Compression without Entropy Coding
11. ICML 2026Efficient Mismatch-Tolerant Coding for Model-Driven Compression
12. ICML 2026End-to-End Compression for Tabular Foundation Models
13. ICML 2026EntroKV: Entropy-Guided Dynamic Budget Allocation for KV-Cache Compression
14. ICML 2026Fast KV Compaction via Attention Matching
15. ICML 2026GradMem: Learning to Write Context into Memory with Test-Time Gradient Descent
16. ICML 2026HARD-KV: Head-Adaptive Regularization for Decoding-time KV Compression
17. ICML 2026How Does the Pretraining Distribution Shape In-Context Learning? A Fundamental Trade-Off
18. ICML 2026How Many Different Outputs Can a Transformer Generate?
19. ICML 2026IndexMem: Learned KV-Cache Eviction with Latent Memory for Long-Context LLM Inference
20. ICML 2026Induction Heads Interpolate N-Grams
21. ICML 2026Less Is More: Elevating RAG via Performance-Driven Context Compression
22. ICML 2026ManifoldKV: Training-Free KV Cache Compression via Euclidean Outlier Detection
23. ICML 2026MaxSAT-Based Compression for Tsetlin Machines
24. ICML 2026NanoSpec: Accelerating Speculative Decoding using Minimalist In-Context Vocabularies
25. ICML 2026No Need to Train Your RDB Foundation Model
26. ICML 2026PACE: Post-Causal Entropy Modeling for Learned LiDAR Point Cloud Compression
27. ICML 2026ParaTool: Shifting Tool Representations from Context to Parameters
28. ICML 2026Predictable Compression Failures: Order Sensitivity and Information Budgeting for Evidence-Grounded Binary Adjudication
29. ICML 2026Principled SVD-based Delta Compression via Quantization Error Minimization
30. ICML 2026Progressive Cramming: Reliable Token Compression and What It Reveals
31. ICML 2026RE-TRAC: REcursive TRAjectory Compression for Deep Search Agents
32. ICML 2026RePo: Language Models with Context Re-Positioning
33. ICML 2026Residual Context Diffusion Language Models
34. ICML 2026Semantic Integrity Matters: Benchmarking and Preserving High-Density Reasoning in KV Cache Compression
35. ICML 2026Softmax as Linear Attention in the Large-Prompt Regime: a Measure-based Perspective
36. ICML 2026Symmetry Reveals Layerwise Dynamics: How Transformers Perform In-Context Classification
37. ICML 2026Transformers Can Learn Posterior Predictive Distributions In-Context
38. ICML 2026Universal Approximation with Softmax Attention
39. ICML 2026xKV: Cross-Layer KV-Cache Compression via Aligned Singular Vector Extraction

常见问题

2023–2026 年四大会议一共有多少篇上下文压缩论文？

共 365 篇，按年份分别是 2023 年 39 篇、2024 年 81 篇、2025 年 109 篇、2026 年 136 篇。

NeurIPS 2026 有多少篇上下文压缩论文？

NeurIPS 2026 有 50 篇上下文压缩论文，上一届是 38 篇。

ICLR 2026 有多少篇上下文压缩论文？

ICLR 2026 有 28 篇上下文压缩论文，上一届是 32 篇。

ICML 2026 有多少篇上下文压缩论文？

ICML 2026 有 39 篇上下文压缩论文，上一届是 20 篇。

CVPR 2026 有多少篇上下文压缩论文？

CVPR 2026 有 19 篇上下文压缩论文，上一届是 19 篇。

哪个会议的上下文压缩论文最多？

NeurIPS，2023–2026 年共 123 篇。