上下文压缩论文趋势 2023–2026
KV 压缩
统计范围：NeurIPS、ICLR、ICML、CVPR 四个会议，2023–2026 年，共 365 篇被归入该主题的论文。
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。
各会议论文数
按年份和会议统计
| 年份 | NeurIPS | ICLR | ICML | CVPR | 合计 |
|---|---|---|---|---|---|
| 2023 | 8 | 13 | 5 | 13 | 39 |
| 2024 | 27 | 15 | 33 | 6 | 81 |
| 2025 | 38 | 32 | 20 | 19 | 109 |
| 2026 | 50 | 28 | 39 | 19 | 136 |
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 365 篇，第 4 / 8 页
- CVPR 2025Random Conditioning for Diffusion Model Compression with Distillation
- CVPR 2025Random Conditioning with Distillation for Data-Efficient Diffusion Model Compression
- CVPR 2025Temporal Action Detection Model Compression by Progressive Block Drop
- CVPR 2025TopNet: Transformer-Efficient Occupancy Prediction Network for Octree-Structured Point Cloud Geometry Compression
- CVPR 2025VoCo-LLaMA: Towards Vision Compression with Large Language Models
- ICLR 2025Approaching Rate-Distortion Limits in Neural Compression with Lattice Transform Coding
- ICLR 2025BenTo: Benchmark Reduction with In-Context Transferability
- ICLR 2025BitStack: Any-Size Compression of Large Language Models in Variable Memory Environments
- ICLR 2025Can In-context Learning Really Generalize to Out-of-distribution Tasks?
- ICLR 2025CAT-3DGS: A Context-Adaptive Triplane Approach to Rate-Distortion-Optimized 3DGS Compression
- ICLR 2025Context-Parametric Inversion: Why Instruction Finetuning May Not Actually Improve Context Reliance
- ICLR 2025DiffPC: Diffusion-based High Perceptual Fidelity Image Compression with Semantic Refinement
- ICLR 2025ICLR: In-Context Learning of Representations
- ICLR 2025Implicit In-context Learning
- ICLR 2025Improving the Sparse Structure Learning of Spiking Neural Networks from the View of Compression Efficiency
- ICLR 2025Learning Randomized Algorithms with Transformers
- ICLR 2025Long Context Compression with Activation Beacon
- ICLR 2025Lossy Compression with Pretrained Diffusion Models
- ICLR 2025MambaExtend: A Training-Free Approach to Improve Long Context Extension of Mamba
- ICLR 2025MCNC: Manifold-Constrained Reparameterization for Neural Compression
- ICLR 2025MELODI: Exploring Memory Compression for Long Contexts
- ICLR 2025Metalic: Meta-Learning In-Context with Protein Language Models
- ICLR 2025MoDeGPT: Modular Decomposition for Large Language Model Compression
- ICLR 2025On the Learn-to-Optimize Capabilities of Transformers in In-Context Sparse Recovery
- ICLR 2025Palu: KV-Cache Compression with Low-Rank Projection
- ICLR 2025PICASO: Permutation-Invariant Context Composition with State Space Models
- ICLR 2025RazorAttention: Efficient KV Cache Compression Through Retrieval Heads
- ICLR 2025ReAttention: Training-Free Infinite Context with Finite Attention Scope
- ICLR 2025Rethinking Invariance in In-context Learning
- ICLR 2025Streamlining Redundant Layers to Compress Large Language Models
- ICLR 2025ThinK: Thinner Key Cache by Query-Driven Pruning
- ICLR 2025Training Free Exponential Context Extension via Cascading KV Cache
- ICLR 2025Transformers are Universal In-context Learners
- ICLR 2025Unsupervised Meta-Learning via In-Context Learning
- ICLR 2025Why Does the Effective Context Length of LLMs Fall Short?
- ICLR 2025Why In-Context Learning Models are Good Few-Shot Learners?
- ICLR 2025You Only Prune Once: Designing Calibration-Free Model Compression With Policy Learning
- ICML 2025CommVQ: Commutative Vector Quantization for KV Cache Compression
- ICML 2025Compression via Pre-trained Transformers: A Study on Byte-Level Multimodal Data
- ICML 2025DLP: Dynamic Layerwise Pruning in Large Language Models
- ICML 2025Does learning the right latent variables necessarily improve in-context learning?
- ICML 2025Efficient LiDAR Reflectance Compression via Scanning Serialization
- ICML 2025Evolving Prompts In-Context: An Open-ended, Self-replicating Perspective
- ICML 2025EvoPress: Accurate Dynamic Model Compression via Evolutionary Search
- ICML 2025In-Context Deep Learning via Transformer Models
- ICML 2025In-Context Learning and Occam's Razor
- ICML 2025Iterative Vectors: In-Context Gradient Steering without Backpropagation
- ICML 2025Learning In-context $n$-grams with Transformers: Sub-$n$-grams Are Near-Stationary Points
- ICML 2025Lexico: Extreme KV Cache Compression via Sparse Coding over Universal Dictionaries
- ICML 2025NExtLong: Toward Effective Long-Context Training without Long Documents
在 Claude、Cursor 里直接问这个趋势
接上 Paperlist MCP 后，写综述、挑方向或准备投稿时，不用离开对话框就能拿到这页的数字：每年各会议的论文数、增长倍数、涨得最快的子方向和代表论文，AI 会按你的问题整理。
可以直接复制这句话问 AI
用 paperlist 查上下文压缩在 NeurIPS 2023–2026 的论文趋势，哪一年涨得最快？
AI 会调用的工具：get_topic_stats {"topic": "compression-context"}
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。