概念解释论文趋势 2023–2026

主要结论

• 概念解释论文从 2023 年的 91 篇增长到 2026 年的 117 篇，三年间扩大到约 1.3 倍。
• NeurIPS 是发概念解释论文最多的会议，2023–2026 年共 134 篇，占四个会议总量的 39%。
• 交叉方向里涨得最快的是多模态大模型：同时属于两个主题的论文从 2023 年的 4 篇增至 2026 年的 23 篇。
• 在「可解释性」的 5 个主题里，它的 2026 年论文数排第 3；同组增长最快的机制解释是 8.1 倍，概念解释是 1.3 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 2026CENDRe: Concept Extraction with Natural Domain Representations
• ICML 2026Evidential Copula Concept Embedding Models
• ICLR 2026Decoupling the Class Label and the Target Concept in Machine Unlearning
• CVPR 2026Agile Deliberation Concept Deliberation for Subjective Visual Classification
• NeurIPS 2025An Analysis of Concept Bottleneck Models: Measuring, Understanding, and Mitigating the Impact of Noisy Annotations
• ICML 2025Concept-Centric Token Interpretation for Vector-Quantized Generative Models
• ICLR 2025Causal Concept Graph Models: Beyond Causal Opacity in Deep Learning
• CVPR 2025Discovering Fine-Grained Visual-Concept Relations by Disentangled Optimal Transport Concept Bottleneck Models
• NeurIPS 2024[Re] On the Reproducibility of Post-Hoc Concept Bottleneck Models
• ICML 2024ContPhy: Continuum Physical Concept Learning and Reasoning from Videos
• ICLR 2024AttEXplore: Attribution for Explanation with model parameters eXploration
• CVPR 2024Fine-Grained Bipartite Concept Factorization for Clustering
• NeurIPS 2023[Re] Hierarchical Shrinkage: Improving the Accuracy and Interpretability of Tree-Based Methods
• ICML 2023A Closer Look at the Intervention Procedure of Concept Bottleneck Models
• ICLR 2023Concept Gradient: Concept-based Interpretation Without Linear Assumption
• CVPR 2023CRAFT: Concept Recursive Activation FacTorization for Explainability
• NeurIPS 2026Concept frustration: Aligning human concepts and machine representations
• ICML 2026On the Variability of Concept Activation Vectors
• ICLR 2026Hierarchical Concept-based Interpretable Models
• CVPR 2026DCoAR Deep Concept Injection into Unified Autoregressive Models for Personalized

全部概念解释论文

共 342 篇，第 3 / 7 页

101. NeurIPS 2026NeuroFaith: Evaluating Mechanistic Faithfulness of LLM Free Text Self-Explanation at the Concept Level
102. NeurIPS 2026NORMA: Norm-Guided Explanation Subgraph Discovery
103. NeurIPS 2026On the Provable Emergence of Hierarchical Concept Structure in CLIP Embeddings
104. NeurIPS 2026PLACE: Patch-Level Agnostic Concept Extraction
105. NeurIPS 2026Position Without Positional Embeddings: A Directional Mechanism in NoPE Transformers
106. NeurIPS 2026Position-aware eXplanation: A Model-Agnostic Framework for Positional Attributions
107. NeurIPS 2026Predictive Concept Decoders: Training Scalable End-to-End Interpretability Assistants
108. NeurIPS 2026Predictive Surprise as Self-Grounding Concept Bottleneck for Interpretable Time Series
109. NeurIPS 2026Second-Order Complexity Theory for Risk, Explanation, and Calibration in Machine Learning
110. NeurIPS 2026Settling the Sample Complexity of Deterministic Agnostic PAC Learning
111. NeurIPS 2026Simultaneous Gradient Learning in First-Price Auctions
112. NeurIPS 2026TanGCE: Manifold-Aware Concept Erasure
113. NeurIPS 2026The Quiet Prompt: Erasing Ineffable Styles from Diffusion Models via Concept Leakage-aware Negative Guidance
114. NeurIPS 2026The SuperActivator Mechanism: Transformers Concentrate Reliable Concept Signals in the Tail
115. NeurIPS 2026TRACER: Token ReAssignment for Concept ERasure in Generative Recommendation
116. NeurIPS 2026Weakly Supervised Concept Learning for Interpreting and Attributing LVLM Predictions
117. NeurIPS 2026You Can’t Have It Both Ways: Concept Entanglement Limits Diffusion Model Unlearning
118. CVPR 2025Attribute-formed Class-specific Concept Space: Endowing Language Bottleneck Model with Better Interpretability and Scalability
119. CVPR 2025CoE: Chain-of-Explanation via Automatic Visual Concept Circuit Description and Polysemanticity Quantification
120. CVPR 2025Concept Replacer: Replacing Sensitive Concepts in Diffusion Models via Precision Localization
121. CVPR 2025Discovering Fine-Grained Visual-Concept Relations by Disentangled Optimal Transport Concept Bottleneck Models
122. CVPR 2025Efficient Fine-Tuning and Concept Suppression for Pruned Diffusion Models
123. CVPR 2025ESC: Erasing Space Concept for Knowledge Deletion
124. CVPR 2025Explaining Domain Shifts in Language: Concept Erasing for Interpretable Image Classification
125. CVPR 2025Finer-CAM: Spotting the Difference Reveals Finer Details for Visual Explanation
126. CVPR 2025Hybrid Concept Bottleneck Models
127. CVPR 2025Interactive Medical Image Analysis with Concept-based Similarity Reasoning
128. CVPR 2025Interpretable Generative Models through Post-hoc Concept Bottlenecks
129. CVPR 2025Language Guided Concept Bottleneck Models for Interpretable Continual Learning
130. CVPR 2025Movie Weaver: Tuning-Free Multi-Concept Video Personalization with Anchored Prompts
131. CVPR 2025Precise, Fast, and Low-cost Concept Erasure in Value Space: Orthogonal Complement Matters
132. CVPR 2025Show and Tell: Visually Explainable Deep Neural Nets via Spatially-Aware Concept Bottleneck Models
133. CVPR 2025Towards Human-Understandable Multi-Dimensional Concept Discovery
134. ICLR 2025Adapting Multi-modal Large Language Model to Concept Drift From Pre-training Onwards
135. ICLR 2025AutoCGP: Closed-Loop Concept-Guided Policies from Unlabeled Demonstrations
136. ICLR 2025Causal Concept Graph Models: Beyond Causal Opacity in Deep Learning
137. ICLR 2025CFD: Learning Generalized Molecular Representation via Concept-Enhanced Feedback Disentanglement
138. ICLR 2025Concept Bottleneck Large Language Models
139. ICLR 2025Concept-ROT: Poisoning Concepts in Large Language Models with Model Editing
140. ICLR 2025ConceptPrune: Concept Editing in Diffusion Models via Skilled Neuron Pruning
141. ICLR 2025Counterfactual Concept Bottleneck Models
142. ICLR 2025Data-centric Prediction Explanation via Kernelized Stein Discrepancy
143. ICLR 2025Efficient and Accurate Explanation Estimation with Distribution Compression
144. ICLR 2025Fantastic Targets for Concept Erasure in Diffusion Models and Where To Find Them
145. ICLR 2025From Tokens to Lattices: Emergent Lattice Structures in Language Models
146. ICLR 2025InstantSwap: Fast Customized Concept Swapping across Sharp Shape Differences
147. ICLR 2025MindSimulator: Exploring Brain Concept Localization via Synthetic fMRI
148. ICLR 2025Navigating Neural Space: Revisiting Concept Activation Vectors to Overcome Directional Divergence
149. ICLR 2025Restyling Unsupervised Concept Based Interpretable Networks with Generative Models
150. ICLR 2025Start Smart: Leveraging Gradients For Enhancing Mask-based XAI Methods

常见问题

2023–2026 年四大会议一共有多少篇概念解释论文？

共 342 篇，按年份分别是 2023 年 91 篇、2024 年 63 篇、2025 年 71 篇、2026 年 117 篇。

NeurIPS 2026 有多少篇概念解释论文？

NeurIPS 2026 有 40 篇概念解释论文，上一届是 25 篇。

ICLR 2026 有多少篇概念解释论文？

ICLR 2026 有 24 篇概念解释论文，上一届是 19 篇。

ICML 2026 有多少篇概念解释论文？

ICML 2026 有 20 篇概念解释论文，上一届是 11 篇。

CVPR 2026 有多少篇概念解释论文？

CVPR 2026 有 33 篇概念解释论文，上一届是 16 篇。

哪个会议的概念解释论文最多？

NeurIPS，2023–2026 年共 134 篇。