特征归因论文趋势 2023–2026

主要结论

• 特征归因论文从 2023 年的 118 篇增长到 2026 年的 219 篇，三年间扩大到约 1.9 倍。
• NeurIPS 是发特征归因论文最多的会议，2023–2026 年共 193 篇，占四个会议总量的 28%。
• 交叉方向里涨得最快的是理论 · learning theory：同时属于两个主题的论文从 2023 年的 1 篇增至 2026 年的 14 篇。
• 在「可解释性」的 5 个主题里，它的 2026 年论文数排第 1；同组增长最快的机制解释是 8.1 倍，特征归因是 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 2026ADAPT: Hybrid Prompt Optimization for LLM Feature Visualization
• ICML 2026Activation with Intrinsic-Extrinsic Consensus
• ICLR 2026AnyUp: Universal Feature Upsampling
• CVPR 2026A Training-Free Style-Personalization via SVD-Based Feature Decomposition
• NeurIPS 2025Alternating Gradient Flows: A Theory of Feature Learning in Two-layer Neural Networks
• ICML 2025BCE vs. CE in Deep Feature Learning
• ICLR 2025A Differentiable Rank-Based Objective for Better Feature Learning
• CVPR 2025A Simple Data Augmentation for Feature Distribution Skewed Federated Learning
• NeurIPS 2024$\texttt{dattri}$: A Library for Efficient Data Attribution
• ICML 2024Asymptotics of feature learning in two-layer networks after one gradient-step
• ICLR 2024An Intuitive Multi-Frequency Feature Representation for SO(3)-Equivariant Networks
• CVPR 2024DYSON: Dynamic Feature Space Self-Organization for Online Task-Free Class Incremental Learning
• NeurIPS 2023$\mathcal{M}^4$: A Unified XAI Benchmark for Faithfulness Evaluation of Feature Attribution Methods across Metrics, Modalities and Models
• ICML 2023Decoding Layer Saliency in Language Transformers
• ICLR 2023Can We Faithfully Represent Absence States to Compute Shapley Values on a DNN?
• CVPR 2023A Practical Upper Bound for the Worst-Case Attribution Deviations
• NeurIPS 2026Adaptive Feature Propagation for Attribute-Missing Graph Clustering
• ICML 2026Aggregate Models, Not Explanations: Improving Feature Importance Estimation
• ICLR 2026Attribution-Guided Decoding
• CVPR 2026FBTA Enabling Single-GPU End-to-End Gigapixel WSI Classification with Feature Bridging

全部特征归因论文

共 190 篇，第 3 / 4 页

101. ICML 2025From Kernels to Features: A Multi-Scale Adaptive Theory of Feature Learning
102. ICML 2025From Pixels to Perception: Interpretable Predictions via Instance-wise Grouped Feature Selection
103. ICML 2025Gamma Distribution PCA-Enhanced Feature Learning for Angle-Robust SAR Target Recognition
104. ICML 2025GEFA: A General Feature Attribution Framework Using Proxy Gradient Estimation
105. ICML 2025Global Convergence and Rich Feature Learning in $L$-Layer Infinite-Width Neural Networks under $\mu$ Parametrization
106. ICML 2025Gradient-based Explanations for Deep Learning Survival Models
107. ICML 2025Identifying and Understanding Cross-Class Features in Adversarial Training
108. ICML 2025Mahalanobis++: Improving OOD Detection via Feature Normalization
109. ICML 2025Mixture of Experts Provably Detect and Learn the Latent Cluster Structure in Gradient-Based Learning
110. ICML 2025No Free Lunch from Random Feature Ensembles: Scaling Laws and Near-Optimality Conditions
111. ICML 2025Nonlinear transformers can perform inference-time feature learning
112. ICML 2025On the Clean Generalization and Robust Overfitting in Adversarial Training from Two Theoretical Views: Representation Complexity and Training Dynamics
113. ICML 2025On The Concurrence of Layer-wise Preconditioning Methods and Provable Feature Learning
114. ICML 2025On the Role of Label Noise in the Feature Learning Process
115. ICML 2025One Wave To Explain Them All: A Unifying Perspective On Feature Attribution
116. ICML 2025One-Pass Feature Evolvable Learning with Theoretical Guarantees
117. ICML 2025Partition First, Embed Later: Laplacian-Based Feature Partitioning for Refined Embedding and Visualization of High-Dimensional Data
118. ICML 2025Propagate and Inject: Revisiting Propagation-Based Feature Imputation for Graphs with Partially Observed Features
119. ICML 2025Random Feature Representation Boosting
120. ICML 2025SAND: One-Shot Feature Selection with Additive Noise Distortion
121. ICML 2025SPEX: Scaling Feature Interaction Explanations for LLMs
122. ICML 2025Statistical Test for Feature Selection Pipelines by Selective Inference
123. ICML 2025Stochastic Encodings for Active Feature Acquisition
124. ICML 2025Stray Intrusive Outliers-Based Feature Selection on Intra-Class Asymmetric Instance Distribution or Multiple High-Density Clusters
125. ICML 2025Sum-of-Parts: Self-Attributing Neural Networks with End-to-End Learning of Feature Groups
126. ICML 2025TabFSBench: Tabular Benchmark for Feature Shifts in Open Environments
127. ICML 2025TAROT: Targeted Data Selection via Optimal Transport
128. ICML 2025The Complexity of Learning Sparse Superposed Features with Feedback
129. ICML 2025The Importance of Being Lazy: Scaling Limits of Continual Learning
130. ICML 2025Theoretical Limitations of Ensembles in the Age of Overparameterization
131. ICML 2025Token Signature: Predicting Chain-of-Thought Gains with Token Decoding Feature in Large Language Models
132. ICML 2025Towards an Explainable Comparison and Alignment of Feature Embeddings
133. ICML 2025Weak-to-Strong Generalization Even in Random Feature Networks, Provably
134. NeurIPS 2025A is for Absorption: Studying Feature Splitting and Absorption in Sparse Autoencoders
135. NeurIPS 2025Accelerating Feature Conformal Prediction via Taylor Approximation
136. NeurIPS 2025Alternating Gradient Flows: A Theory of Feature Learning in Two-layer Neural Networks
137. NeurIPS 2025Approximating Shapley Explanations in Reinforcement Learning
138. NeurIPS 2025Attribution-Driven Adaptive Token Pruning for Transformers
139. NeurIPS 2025Better Training Data Attribution via Better Inverse Hessian-Vector Products
140. NeurIPS 2025BevSplat: Resolving Height Ambiguity via Feature-Based Gaussian Primitives for Weakly-Supervised Cross-View Localization
141. NeurIPS 2025Beyond Last-Click: An Optimal Mechanism for Ad Attribution
142. NeurIPS 2025Capturing Polysemanticity with PRISM: A Multi-Concept Feature Description Framework
143. NeurIPS 2025Closed-Form Training Dynamics Reveal Learned Features and Linear Structure in Word2Vec-like Models
144. NeurIPS 2025Contrastive Learning with Data Misalignment: Feature Purity, Training Dynamics and Theoretical Generalization Guarantees
145. NeurIPS 2025DATE-LM: Benchmarking Data Attribution Evaluation for Large Language Models
146. NeurIPS 2025Degrees of Freedom for Linear Attention: Distilling Softmax Attention with Optimal Feature Efficiency
147. NeurIPS 2025Demystifying Spectral Feature Learning for Instrumental Variable Regression
148. NeurIPS 2025DePass: Unified Feature Attributing by Simple Decomposed Forward Pass
149. NeurIPS 2025Distributional Training Data Attribution: What do Influence Functions Sample?
150. NeurIPS 2025Distributionally Robust Feature Selection

常见问题

2023–2026 年四大会议一共有多少篇特征归因论文？

共 695 篇，按年份分别是 2023 年 118 篇、2024 年 168 篇、2025 年 190 篇、2026 年 219 篇。

NeurIPS 2026 有多少篇特征归因论文？

NeurIPS 2026 有 61 篇特征归因论文，上一届是 57 篇。

ICLR 2026 有多少篇特征归因论文？

ICLR 2026 有 53 篇特征归因论文，上一届是 42 篇。

ICML 2026 有多少篇特征归因论文？

ICML 2026 有 56 篇特征归因论文，上一届是 51 篇。

CVPR 2026 有多少篇特征归因论文？

CVPR 2026 有 49 篇特征归因论文，上一届是 40 篇。

哪个会议的特征归因论文最多？

NeurIPS，2023–2026 年共 193 篇。