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 117 篇，第 2 / 3 页

51. ICLR 2026Tackling Time-Series Forecasting Generalization via Mitigating Concept Drift
52. ICLR 2026Temporal Concept Dynamics in Diffusion Models via Prompt-Conditioned Interventions
53. ICLR 2026The Lattice Representation Hypothesis and Concept Algebra in LLMs
54. ICLR 2026There Was Never a Bottleneck in Concept Bottleneck Models
55. ICLR 2026Training-free Counterfactual Explanation for Temporal Graph Model Inference
56. ICLR 2026When Machine Learning Gets Personal: Evaluating Prediction and Explanation
57. ICLR 2026When to Retrain after Drift: A Data-Only Test of Post-Drift Data Size Sufficiency
58. ICML 2026ConEx: Human-Interpretable Saliency Maps via Concept-Aware Attribution
59. ICML 2026Erased but Not Forgotten: How Backdoors Compromise Concept Erasure
60. ICML 2026Evidential Copula Concept Embedding Models
61. ICML 2026Explaining Concept Shift with Interpretable Feature Attribution
62. ICML 2026Explanations are a Means to an End: Decision Theoretic Explanation Evaluation
63. ICML 2026Formal Concept Lattices are Good Semantic Scaffolds for Concept-Based Learning
64. ICML 2026General Quantification of Covariate and Concept Shifts
65. ICML 2026Gradient Testing and Estimation by Comparisons
66. ICML 2026Learning Partial Concept Classes and Universal Rates Under Massart Noise
67. ICML 2026Mixture of Concept Bottleneck Experts
68. ICML 2026MoSSP: A Momentum-Based Single-Loop Stochastic Penalty Method for Nonconvex Constrained DC-regularized Optimization
69. ICML 2026Neural Concept Verifier: Scaling Prover-Verifier Games via Concept Encodings
70. ICML 2026Neuro-Fuzzy Concept Learning for Interpretable Large Multimodal Models
71. ICML 2026On Learnability and Disambiguation of Multiclass Partial Concept Classes
72. ICML 2026On the Variability of Concept Activation Vectors
73. ICML 2026Polaris: Coupled Orbital Polar Embeddings for Hierarchical Concept Learning
74. ICML 2026PPDL: LLM-Based Flows as Probabilistic Programs
75. ICML 2026Prototype-Grounded Concept Models for Verifiable Concept Alignment
76. ICML 2026RC-FCL: Combating Asynchronous Concept Drift in Federated Continual Learning via Retrospective Calibration
77. ICML 2026What LLMs Explain Is Not What They Believe: Evaluating Explanation Sufficiency Under Models' Own Input Beliefs
78. NeurIPS 2026A Theoretical Analysis of Why Masked Diffusion Models Mitigate the Reversal Curse
79. NeurIPS 2026CENDRe: Concept Extraction with Natural Domain Representations
80. NeurIPS 2026Concept frustration: Aligning human concepts and machine representations
81. NeurIPS 2026Concepts Worth Having: Refining VLM-Guided Concept Bottleneck Models with Minimal Annotations
82. NeurIPS 2026Concise and Logically Consistent Conformal Sets for Neuro-Symbolic Concept-Based Models
83. NeurIPS 2026Differential Vector Erasure: Unified Training-Free Concept Erasure for Flow Matching Models
84. NeurIPS 2026Diffusion-Time Concept Manifolds: Sparse Autoencoder Groups for Interpreting Denoising Language Models
85. NeurIPS 2026Distribution Free Fourier Sparsity Testing
86. NeurIPS 2026Do Sparse Autoencoders Learn Meaningful Concept Hierarchies?
87. NeurIPS 2026Do You CARE to Generalize? Extracting Robust Concept Directions from LLMs
88. NeurIPS 2026Example-Based Spatial Guidance for Training-Free Concept Erasure in Diffusion Models
89. NeurIPS 2026Explanation Multiplicity in SHAP: Characterization and Assessment
90. NeurIPS 2026Explanation of Dynamic Physical Field Predictions using WassersteinGrad: Application to Autoregressive Weather Forecasting
91. NeurIPS 2026Few-shot Task Learning via Compositional Concept Inference
92. NeurIPS 2026From Zero to Hero: Training-Free Custom Concept Spawning in World Models
93. NeurIPS 2026Hierarchical Concept Geometry in Language Representations Emerges from Word Co-occurrence
94. NeurIPS 2026Hyperbolic Concept Bottleneck Models
95. NeurIPS 2026Introspective Coupling: LMs Learn to Explain Themselves Better Than Their Training Targets
96. NeurIPS 2026Latent Introspection: Models Can Detect Prior Concept Injections
97. NeurIPS 2026Lost or Hidden? A Concept-Level Forgetting in Supervised Continual Learning
98. NeurIPS 2026Marginal-Nonuniform Multiclass Learning
99. NeurIPS 2026Measuring What Matters: Synthetic Benchmarks for Concept Bottleneck Models
100. NeurIPS 2026MOSAIC: Concept Bottlenecks via Text-Anchored Optimal Transport

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。