稠密检索论文趋势 2023–2026

主要结论

• 稠密检索论文从 2023 年的 60 篇增长到 2026 年的 162 篇，三年间扩大到约 2.7 倍。
• CVPR 是发稠密检索论文最多的会议，2023–2026 年共 147 篇，占四个会议总量的 37%。
• 交叉方向里涨得最快的是检索记忆：同时属于两个主题的论文从 2023 年的 1 篇增至 2026 年的 21 篇。
• 在「信息检索」的 5 个主题里，它的 2026 年论文数排第 1；同组增长最快的交叉编码重排是 3.0 倍，稠密检索是 2.7 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 2026Half-Truths Break Similarity-Based Retrieval
• ICML 2026Deep Discriminative Structure Proxy Hashing for Cross-modal Retrieval
• ICLR 2026A Dense Subset Index for Collective Query Coverage
• CVPR 2026Beyond Caption-Based Queries in Video Moment Retrieval
• NeurIPS 2025Dense Associative Memory with Epanechnikov Energy
• ICML 2025Nonparametric Modern Hopfield Models
• ICLR 2025Contextual Document Embeddings
• CVPR 2025Adapting Dense Matching for Homography Estimation with Grid-based Acceleration
• NeurIPS 2024CARE: a Benchmark Suite for the Classification and Retrieval of Enzymes
• ICML 2024Compute Better Spent: Replacing Dense Layers with Structured Matrices
• ICLR 2024Adaptive Retrieval and Scalable Indexing for k-NN Search with Cross-Encoders
• CVPR 2024Dense Optical Tracking: Connecting the Dots
• NeurIPS 2023Bypassing spike sorting: Density-based decoding using spike localization from dense multielectrode probes
• ICML 2023Everyone's Preference Changes Differently: A Weighted Multi-Interest Model For Retrieval
• ICLR 2023Automating Nearest Neighbor Search Configuration with Constrained Optimization
• CVPR 2023Are Binary Annotations Sufficient? Video Moment Retrieval via Hierarchical Uncertainty-Based Active Learning
• NeurIPS 2026Hyperparameter Transfer for Dense Associative Memories
• ICML 2026Dense associative memory for Gaussian distributions
• ICLR 2026Beyond Instance-Level Alignment: Dual-Level Optimal Transport for Audio-Text Retrieval
• CVPR 2026Beyond Tie Points Satellite Image Block Adjustment based on Dense

全部稠密检索论文

共 401 篇，第 3 / 9 页

101. ICML 2026PCA of Probability Measures: Sparse and Dense Sampling Regimes
102. ICML 2026RAT+: Train Dense, Infer Sparse - Recurrence Augmented Attention for Dilated Inference
103. ICML 2026Robust Self-reflective Hashing for Cross-modal Retrieval with Noisy Label
104. ICML 2026Similarity Is Not Logic: Factored Inference for Dual-Encoder Vision-Language Models
105. ICML 2026SOLAR: Self-supervised Joint Learning for Symmetric Multimodal Retrieval
106. ICML 2026Spectral Gradient Descent Mitigates Anisotropy-Driven Misalignment: A Case Study in Phase Retrieval
107. ICML 2026Temporal Preference Optimization for Unsupervised Retrieval
108. ICML 2026The Crowded Embedding Space: A Mean-Field Mechanism for Emergent Marginalization in Retrieval-Augmented Agents
109. ICML 2026TimeRewarder: Learning Dense Reward from Passive Videos via Frame-wise Temporal Distance
110. ICML 2026Unifying Adversarial Robustness and Training Across Text Scoring Models
111. ICML 2026With Argus Eyes: Assessing Retrieval Gaps via Uncertainty Scoring to Detect and Remedy Retrieval Blind Spots
112. NeurIPS 2026AC/DC on a Budget -- Alternating Sparse Phases
113. NeurIPS 2026Adversarial Corpus Selection to Attack Subgraph Matching based Graph Retrieval
114. NeurIPS 2026Beyond RAG for Agent Memory: Retrieval by Decoupling and Aggregation
115. NeurIPS 2026Beyond Semantic Similarity: Rethinking Retrieval for Agentic Search via Direct Corpus Interaction
116. NeurIPS 2026Beyond Spatial and Temporal Priors: A Generalizable Approach for Dense Correspondence Matching
117. NeurIPS 2026Can Language Models Actually Retrieve In-Context? Drowning in Documents at Million Token Scale
118. NeurIPS 2026Causal-VLM: Dense Causal Captioning in Videos
119. NeurIPS 2026Chain of Dual Structures in Transformer Attention
120. NeurIPS 2026Constraint Retrieval Is Not Constraint Enforcement in Large Language Models
121. NeurIPS 2026Contrastive Retrieval Heads for Improved Attention-Based Reranking
122. NeurIPS 2026CyberDualEval: Measuring Dual-Use Cyber Risks in Frontier Language Models
123. NeurIPS 2026Dense Cross-Tokenizer Distillation via Semantic Optimal Local Alignment
124. NeurIPS 2026Dual-Rate Diffusion: Accelerating diffusion models with an interleaved heavy-light network
125. NeurIPS 2026FlashSSM-3D: A Distilled State Space Model for Lightning-Fast Dense 3D Reconstruction
126. NeurIPS 2026Group of Skills: Group-Structured Skill Retrieval for Agent Skill Libraries
127. NeurIPS 2026Half-Truths Break Similarity-Based Retrieval
128. NeurIPS 2026How Much Evidence Should Retrieval-Augmented In-Context Learning Use Under Distribution Shift?
129. NeurIPS 2026Hyperparameter Transfer for Dense Associative Memories
130. NeurIPS 2026Is Dimensionality a Barrier for Retrieval Models?
131. NeurIPS 2026Latent Abstraction for Retrieval-Augmented Generation
132. NeurIPS 2026LeanSearch v2: Global Premise Retrieval for Lean 4 Theorem Proving
133. NeurIPS 2026Learning Semantic Consistency for Open-Vocabulary Dense Perception
134. NeurIPS 2026Logit-Contribution Scoring Identifies Non-Literal Retrieval Heads
135. NeurIPS 2026LogSTOP: Temporal Scores over Prediction Sequences for Matching and Retrieval
136. NeurIPS 2026Memory Retrieval for Changing Preferences
137. NeurIPS 2026MLAIRE: Multilingual Language-Aware Information Retrieval Evaluation Protocol
138. NeurIPS 2026Neural Networks With Dense Weights Are Not Universal Approximators
139. NeurIPS 2026On objective mismatch in molecular retrieval from tandem mass spectrometry
140. NeurIPS 2026Optimistic Dual Averaging Unifies Modern Optimizers
141. NeurIPS 2026PANDA: Prior-guided Attentional Dual-path Architecture
142. NeurIPS 2026Phase Kernel Lifts Capacity of Dense Associative Memory
143. NeurIPS 2026PixelDense: Dense Prediction as Representation Alignment for Pixel Diffusion
144. NeurIPS 2026Point Clustering Encoders
145. NeurIPS 2026Precision As You Need: Stochastic Computing Is a Dense Adaptive Quantizer
146. NeurIPS 2026Process-conditioned Pretraining with Topographic Spatial Retrieval for Large EEG Models
147. NeurIPS 2026ProSearch: Benchmarking Multi-Constraint Protocol Retrieval in Experimental Science
148. NeurIPS 2026QB-Highlights: Quality-Guided Budgeted Highlight Detection in Videos with Dense Query-Relevant Moments
149. NeurIPS 2026RAGBoost: Robust Tabular Learning via Retrieval-Augmented and Ancillary-Guided Gradient Boosting
150. NeurIPS 2026Retrieval Over Training: Similarity search-based Model Selection for Time Series Anomaly Detection

常见问题

2023–2026 年四大会议一共有多少篇稠密检索论文？

共 401 篇，按年份分别是 2023 年 60 篇、2024 年 92 篇、2025 年 87 篇、2026 年 162 篇。

NeurIPS 2026 有多少篇稠密检索论文？

NeurIPS 2026 有 51 篇稠密检索论文，上一届是 35 篇。

ICLR 2026 有多少篇稠密检索论文？

ICLR 2026 有 26 篇稠密检索论文，上一届是 12 篇。

ICML 2026 有多少篇稠密检索论文？

ICML 2026 有 35 篇稠密检索论文，上一届是 10 篇。

CVPR 2026 有多少篇稠密检索论文？

CVPR 2026 有 50 篇稠密检索论文，上一届是 30 篇。

哪个会议的稠密检索论文最多？

CVPR，2023–2026 年共 147 篇。