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 123 篇，第 2 / 3 页

51. NeurIPS 2025AdmTree: Compressing Lengthy Context with Adaptive Semantic Trees
52. NeurIPS 2025AttentionPredictor: Temporal Patterns Matter for KV Cache Compression
53. NeurIPS 2025Binary Quadratic Quantization: Beyond First-Order Quantization for Real-Valued Matrix Compression
54. NeurIPS 2025Breaking the Compression Ceiling: Data-Free Pipeline for Ultra-Efficient Delta Compression
55. NeurIPS 2025ChunkKV: Semantic-Preserving KV Cache Compression for Efficient Long-Context LLM Inference
56. NeurIPS 2025COALA: Numerically Stable and Efficient Framework for Context-Aware Low-Rank Approximation
57. NeurIPS 2025Conditional Distribution Compression via the Kernel Conditional Mean Embedding
58. NeurIPS 2025Constant Bit-size Transformers Are Turing Complete
59. NeurIPS 2025DeepHalo: A Neural Choice Model with Controllable Context Effects
60. NeurIPS 2025Disentangling Latent Shifts of In-Context Learning with Weak Supervision
61. NeurIPS 2025Don't Just Chase “Highlighted Tokens” in MLLMs: Revisiting Visual Holistic Context Retention
62. NeurIPS 2025DUO: No Compromise to Accuracy Degradation
63. NeurIPS 2025Efficient Prompt Compression with Evaluator Heads for Long-Context Transformer Inference
64. NeurIPS 2025End-to-End Low-Light Enhancement for Object Detection with Learned Metadata from RAWs
65. NeurIPS 2025FP4 All the Way: Fully Quantized Training of Large Language Models
66. NeurIPS 2025From Softmax to Score: Transformers Can Effectively Implement In-Context Denoising Steps
67. NeurIPS 2025Homogeneous Keys, Heterogeneous Values: Exploiting Local KV Cache Asymmetry for Long-Context LLMs
68. NeurIPS 2025In-Context Compositional Learning vis Sparse Coding Transformer
69. NeurIPS 2025InfiniPot-V: Memory-Constrained KV Cache Compression for Streaming Video Understanding
70. NeurIPS 2025KeyDiff: Key Similarity-Based KV Cache Eviction for Long-Context LLM Inference in Resource-Constrained Environments
71. NeurIPS 2025KVzip: Query-Agnostic KV Cache Compression with Context Reconstruction
72. NeurIPS 2025Language Models (Mostly) Know When to Stop Reading
73. NeurIPS 2025Large Language Models for Lossless Image Compression: Next-Pixel Prediction in Language Space is All You Need
74. NeurIPS 2025LittleBit: Ultra Low-Bit Quantization via Latent Factorization
75. NeurIPS 2025MoRIC: A Modular Region-based Implicit Codec for Image Compression
76. NeurIPS 2025Neural Attention Search
77. NeurIPS 2025Neural B-frame Video Compression with Bi-directional Reference Harmonization
78. NeurIPS 2025Overcoming Long Context Limitations of State Space Models via Context Dependent Sparse Attention
79. NeurIPS 2025Scale-invariant attention
80. NeurIPS 2025SmallKV: Small Model Assisted Compensation of KV Cache Compression for Efficient LLM Inference
81. NeurIPS 2025Stochastic Forward-Forward Learning through Representational Dimensionality Compression
82. NeurIPS 2025Theoretical Insights into In-context Learning with Unlabeled Data
83. NeurIPS 2025UniGist: Towards General and Hardware-aligned Sequence-level Long Context Compression
84. NeurIPS 2025Unlabeled Data Can Provably Enhance In-Context Learning of Transformers
85. NeurIPS 2025Value-Guided KV Compression for LLMs via Approximated CUR Decomposition
86. NeurIPS 2025Vision-centric Token Compression in Large Language Model
87. NeurIPS 2025Vocabulary In-Context Learning in Transformers: Benefits of Positional Encoding
88. NeurIPS 2025zip2zip: Inference-Time Adaptive Tokenization via Online Compression
89. NeurIPS 2024Base of RoPE Bounds Context Length
90. NeurIPS 2024Causal Context Adjustment Loss for Learned Image Compression
91. NeurIPS 2024ContextGS : Compact 3D Gaussian Splatting with Anchor Level Context Model
92. NeurIPS 2024DiTFastAttn: Attention Compression for Diffusion Transformer Models
93. NeurIPS 2024DuQuant: Distributing Outliers via Dual Transformation Makes Stronger Quantized LLMs
94. NeurIPS 2024Efficient Large Multi-modal Models via Visual Context Compression
95. NeurIPS 2024Enhancing In-Context Learning Performance with just SVD-Based Weight Pruning: A Theoretical Perspective
96. NeurIPS 2024ESPACE: Dimensionality Reduction of Activations for Model Compression
97. NeurIPS 2024Euclidean distance compression via deep random features
98. NeurIPS 2024Exploring Context Window of Large Language Models via Decomposed Positional Vectors
99. NeurIPS 2024In-Context Learning with Transformers: Softmax Attention Adapts to Function Lipschitzness
100. NeurIPS 2024KVQuant: Towards 10 Million Context Length LLM Inference with KV Cache Quantization

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。