信息抽取论文趋势 2023–2026

主要结论

• 信息抽取论文从 2023 年的 143 篇增长到 2026 年的 266 篇，三年间扩大到约 1.9 倍。
• NeurIPS 是发信息抽取论文最多的会议，2023–2026 年共 287 篇，占四个会议总量的 36%。
• 交叉方向里涨得最快的是扩散模型：同时属于两个主题的论文从 2023 年的 2 篇增至 2026 年的 28 篇。
• 在「自然语言」的 5 个主题里，它的 2026 年论文数排第 3；同组增长最快的检索增强生成是 5.4 倍，信息抽取是 1.9 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 2026$\texttt{bispectrum}$: Selective $G$-Bispectra Made Practical
• ICML 2026AREA: Attribute Extraction and Aggregation for CLIP-Based Class-Incremental Learning
• ICLR 2026An Information-Theoretic Parameter-Free Bayesian Framework for Probing Labeled Dependency Trees from Attention Score
• CVPR 2026ActiveGrasp Information-Guided Active Grasping with Calibrated Energy-based Model
• NeurIPS 2025A Pre-training Framework for Relational Data with Information-theoretic Principles
• ICML 2025A Peer-review Look on Multi-modal Clustering: An Information Bottleneck Realization Method
• ICLR 2025ADAM: An Embodied Causal Agent in Open-World Environments
• CVPR 2025A Focused Human Body Model for Accurate Anthropometric Measurements Extraction
• NeurIPS 2024TACT: Advancing Complex Aggregative Reasoning with Information Extraction Tools
• ICML 2024An Information-Theoretic Analysis of In-Context Learning
• ICLR 2024Bayes Conditional Distribution Estimation for Knowledge Distillation Based on Conditional Mutual Information
• CVPR 2024Differentiable Information Bottleneck for Deterministic Multi-view Clustering
• NeurIPS 2023A Novel Approach for Effective Multi-View Clustering with Information-Theoretic Perspective
• ICML 2023Abstracting Imperfect Information Away from Two-Player Zero-Sum Games
• ICLR 2023A VAE for Transformers with Nonparametric Variational Information Bottleneck
• CVPR 2023"Seeing" Electric Network Frequency From Events
• NeurIPS 2026An Information-theoretic Framework for Auditing Unfairness in Training Data
• ICML 2026GCIB: Graph Contrastive Information Bottleneck for Multi-Behavior Recommendation
• ICLR 2026Calibrated Information Bottleneck for Trusted Multi-modal Clustering
• CVPR 2026Beyond Whats Shared Recovering Lost Unique Information from Intermediate Layers

全部信息抽取论文

共 199 篇，第 3 / 4 页

101. ICML 2025Learning Likelihood-Free Reference Priors
102. ICML 2025Learning Optimal Multimodal Information Bottleneck Representations
103. ICML 2025Matrix Completion with Incomplete Side Information via Orthogonal Complement Projection
104. ICML 2025NBDI: A Simple and Effective Termination Condition for Skill Extraction from Task-Agnostic Demonstrations
105. ICML 2025Noise-Guided Predicate Representation Extraction and Diffusion-Enhanced Discretization for Scene Graph Generation
106. ICML 2025On the Importance of Gaussianizing Representations
107. ICML 2025Online Curvature-Aware Replay: Leveraging $\mathbf{2^{nd}}$ Order Information for Online Continual Learning
108. ICML 2025Optimal Information Retention for Time-Series Explanations
109. ICML 2025PIGDreamer: Privileged Information Guided World Models for Safe Partially Observable Reinforcement Learning
110. ICML 2025Pointwise Information Measures as Confidence Estimators in Deep Neural Networks: A Comparative Study
111. ICML 2025Position: Language model developers should report train-test overlap
112. ICML 2025ProDiff: Prototype-Guided Diffusion for Minimal Information Trajectory Imputation
113. ICML 2025Revealing Weaknesses in Text Watermarking Through Self-Information Rewrite Attacks
114. ICML 2025ROME is Forged in Adversity: Robust Distilled Datasets via Information Bottleneck
115. ICML 2025Scalable Attribute-Missing Graph Clustering via Neighborhood Differentiation
116. ICML 2025Sharp Optimality of Simple, Plug-in Estimation of the Fisher Information of a Smoothed Density
117. ICML 2025Super Deep Contrastive Information Bottleneck for Multi-modal Clustering
118. ICML 2025The Sample Complexity of Online Strategic Decision Making with Information Asymmetry and Knowledge Transportability
119. ICML 2025ZeroFlow: Overcoming Catastrophic Forgetting is Easier than You Think
120. NeurIPS 2025A Pre-training Framework for Relational Data with Information-theoretic Principles
121. NeurIPS 2025A Semantic Parsing Framework for End-to-End Time Normalization
122. NeurIPS 2025Balanced Active Inference
123. NeurIPS 2025ChemX: A Collection of Chemistry Datasets for Benchmarking Automated Information Extraction
124. NeurIPS 2025ClinBench: A Standardized Multi-Domain Framework for Evaluating Large Language Models in Clinical Information Extraction
125. NeurIPS 2025Collective Bargaining in the Information Economy Can Address AI-Driven Power Concentration
126. NeurIPS 2025Conformal Information Pursuit for Interactively Guiding Large Language Models
127. NeurIPS 2025Connecting Jensen–Shannon and Kullback–Leibler Divergences: A New Bound for Representation Learning
128. NeurIPS 2025CoT Information: Improved Sample Complexity under Chain-of-Thought Supervision
129. NeurIPS 2025CyIN: Cyclic Informative Latent Space for Bridging Complete and Incomplete Multimodal Learning
130. NeurIPS 2025DataSIR: A Benchmark Dataset for Sensitive Information Recognition
131. NeurIPS 2025Decomposing Interventional Causality into Synergistic, Redundant, and Unique Components
132. NeurIPS 2025Decomposing stimulus-specific sensory neural information via diffusion models
133. NeurIPS 2025Demystifying Reasoning Dynamics with Mutual Information: Thinking Tokens are Information Peaks in LLM Reasoning
134. NeurIPS 2025Efficient Kernelized Learning in Polyhedral Games beyond Full Information: From Colonel Blotto to Congestion Games
135. NeurIPS 2025Enhancing Privacy in Multimodal Federated Learning with Information Theory
136. NeurIPS 2025Evolution of Information in Interactive Decision Making: A Case Study for Multi-Armed Bandits
137. NeurIPS 2025Exploiting Vocabulary Frequency Imbalance in Language Model Pre-training
138. NeurIPS 2025Fantastic Features and Where to Find Them: A Probing Method to combine Features from Multiple Foundation Models
139. NeurIPS 2025FedGPS: Statistical Rectification Against Data Heterogeneity in Federated Learning
140. NeurIPS 2025Feedback-Aware MCTS for Goal-Oriented Information Seeking
141. NeurIPS 2025FlowCut: Rethinking Redundancy via Information Flow for Efficient Vision-Language Models
142. NeurIPS 2025FreeControl: Efficient, Training-Free Structural Control via One-Step Attention Extraction
143. NeurIPS 2025From Counterfactuals to Trees: Competitive Analysis of Model Extraction Attacks
144. NeurIPS 2025From Information to Generative Exponent: Learning Rate Induces Phase Transitions in SGD
145. NeurIPS 2025Glocal Information Bottleneck for Time Series Imputation
146. NeurIPS 2025Graph–Smoothed Bayesian Black-Box Shift Estimator and Its Information Geometry
147. NeurIPS 2025Grounding Language with Vision: A Conditional Mutual Information Calibrated Decoding Strategy for Reducing Hallucinations in LVLMs
148. NeurIPS 2025H-SPLID: HSIC-based Saliency Preserving Latent Information Decomposition
149. NeurIPS 2025HawkBench: Investigating Resilience of RAG Methods on Stratified Information-Seeking Tasks
150. NeurIPS 2025Hierarchical Information Aggregation for Incomplete Multimodal Alzheimer's Disease Diagnosis

常见问题

2023–2026 年四大会议一共有多少篇信息抽取论文？

共 798 篇，按年份分别是 2023 年 143 篇、2024 年 190 篇、2025 年 199 篇、2026 年 266 篇。

NeurIPS 2026 有多少篇信息抽取论文？

NeurIPS 2026 有 96 篇信息抽取论文，上一届是 80 篇。

ICLR 2026 有多少篇信息抽取论文？

ICLR 2026 有 63 篇信息抽取论文，上一届是 50 篇。

ICML 2026 有多少篇信息抽取论文？

ICML 2026 有 74 篇信息抽取论文，上一届是 46 篇。

CVPR 2026 有多少篇信息抽取论文？

CVPR 2026 有 33 篇信息抽取论文，上一届是 23 篇。

哪个会议的信息抽取论文最多？

NeurIPS，2023–2026 年共 287 篇。