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 71 篇，第 1 / 2 页

1. CVPR 2025Attribute-formed Class-specific Concept Space: Endowing Language Bottleneck Model with Better Interpretability and Scalability
2. CVPR 2025CoE: Chain-of-Explanation via Automatic Visual Concept Circuit Description and Polysemanticity Quantification
3. CVPR 2025Concept Replacer: Replacing Sensitive Concepts in Diffusion Models via Precision Localization
4. CVPR 2025Discovering Fine-Grained Visual-Concept Relations by Disentangled Optimal Transport Concept Bottleneck Models
5. CVPR 2025Efficient Fine-Tuning and Concept Suppression for Pruned Diffusion Models
6. CVPR 2025ESC: Erasing Space Concept for Knowledge Deletion
7. CVPR 2025Explaining Domain Shifts in Language: Concept Erasing for Interpretable Image Classification
8. CVPR 2025Finer-CAM: Spotting the Difference Reveals Finer Details for Visual Explanation
9. CVPR 2025Hybrid Concept Bottleneck Models
10. CVPR 2025Interactive Medical Image Analysis with Concept-based Similarity Reasoning
11. CVPR 2025Interpretable Generative Models through Post-hoc Concept Bottlenecks
12. CVPR 2025Language Guided Concept Bottleneck Models for Interpretable Continual Learning
13. CVPR 2025Movie Weaver: Tuning-Free Multi-Concept Video Personalization with Anchored Prompts
14. CVPR 2025Precise, Fast, and Low-cost Concept Erasure in Value Space: Orthogonal Complement Matters
15. CVPR 2025Show and Tell: Visually Explainable Deep Neural Nets via Spatially-Aware Concept Bottleneck Models
16. CVPR 2025Towards Human-Understandable Multi-Dimensional Concept Discovery
17. ICLR 2025Adapting Multi-modal Large Language Model to Concept Drift From Pre-training Onwards
18. ICLR 2025AutoCGP: Closed-Loop Concept-Guided Policies from Unlabeled Demonstrations
19. ICLR 2025Causal Concept Graph Models: Beyond Causal Opacity in Deep Learning
20. ICLR 2025CFD: Learning Generalized Molecular Representation via Concept-Enhanced Feedback Disentanglement
21. ICLR 2025Concept Bottleneck Large Language Models
22. ICLR 2025Concept-ROT: Poisoning Concepts in Large Language Models with Model Editing
23. ICLR 2025ConceptPrune: Concept Editing in Diffusion Models via Skilled Neuron Pruning
24. ICLR 2025Counterfactual Concept Bottleneck Models
25. ICLR 2025Data-centric Prediction Explanation via Kernelized Stein Discrepancy
26. ICLR 2025Efficient and Accurate Explanation Estimation with Distribution Compression
27. ICLR 2025Fantastic Targets for Concept Erasure in Diffusion Models and Where To Find Them
28. ICLR 2025From Tokens to Lattices: Emergent Lattice Structures in Language Models
29. ICLR 2025InstantSwap: Fast Customized Concept Swapping across Sharp Shape Differences
30. ICLR 2025MindSimulator: Exploring Brain Concept Localization via Synthetic fMRI
31. ICLR 2025Navigating Neural Space: Revisiting Concept Activation Vectors to Overcome Directional Divergence
32. ICLR 2025Restyling Unsupervised Concept Based Interpretable Networks with Generative Models
33. ICLR 2025Start Smart: Leveraging Gradients For Enhancing Mask-based XAI Methods
34. ICLR 2025Swing-by Dynamics in Concept Learning and Compositional Generalization
35. ICLR 2025XAIguiFormer: explainable artificial intelligence guided transformer for brain disorder identification
36. ICML 2025Avoiding Leakage Poisoning: Concept Interventions Under Distribution Shifts
37. ICML 2025Concept-Centric Token Interpretation for Vector-Quantized Generative Models
38. ICML 2025Deep Streaming View Clustering
39. ICML 2025Editable Concept Bottleneck Models
40. ICML 2025Enhancing Performance of Explainable AI Models with Constrained Concept Refinement
41. ICML 2025FastCAV: Efficient Computation of Concept Activation Vectors for Explaining Deep Neural Networks
42. ICML 2025Nonparametric Identification of Latent Concepts
43. ICML 2025Representation Preserving Multiclass Agnostic to Realizable Reduction
44. ICML 2025TIMING: Temporality-Aware Integrated Gradients for Time Series Explanation
45. ICML 2025Towards a Mechanistic Explanation of Diffusion Model Generalization
46. ICML 2025Universal Sparse Autoencoders: Interpretable Cross-Model Concept Alignment
47. NeurIPS 2025Advancing Interpretability of CLIP Representations with Concept Surrogate Model
48. NeurIPS 2025An Analysis of Concept Bottleneck Models: Measuring, Understanding, and Mitigating the Impact of Noisy Annotations
49. NeurIPS 2025Bayesian Concept Bottleneck Models with LLM Priors
50. NeurIPS 2025Causally Reliable Concept Bottleneck Models

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。