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 53 篇，第 1 / 2 页

1. ICLR 2026$\boldsymbol{\partial^\infty}$-Grid: A Neural Differential Equation Solver with Differentiable Feature Grids
2. ICLR 2026$\mathbf{Li_2}$: A Framework on Dynamics of Feature Emergence and Delayed Generalization
3. ICLR 2026ACE: Attribution-Controlled Knowledge Editing for Multi-hop Factual Recall
4. ICLR 2026Adaptive Regularization for Large-Scale Sparse Feature Embedding Models
5. ICLR 2026AnyUp: Universal Feature Upsampling
6. ICLR 2026Attribution-Guided Decoding
7. ICLR 2026Bayesian Influence Functions for Hessian-Free Data Attribution
8. ICLR 2026Bridging Input Feature Spaces Towards Graph Foundation Models
9. ICLR 2026COMPASS: Robust Feature Conformal Prediction for Medical Segmentation Metrics
10. ICLR 2026Concept-TRAK: Understanding how diffusion models learn concepts through concept attribution
11. ICLR 2026Correlations in the Data Lead to Semantically Rich Feature Geometry Under Superposition
12. ICLR 2026Dissecting Representation Misalignment in Contrastive Learning via Influence Function
13. ICLR 2026DIVERSE: Disagreement-Inducing Vector Evolution for Rashomon Set Exploration
14. ICLR 2026Evaluating Data Influence in Meta Learning
15. ICLR 2026Evolution of Concepts in Language Model Pre-Training
16. ICLR 2026Expressive yet Efficient Feature Expansion with Adaptive Cross-Hadamard Products
17. ICLR 2026f-INE: A Hypothesis Testing Framework for Estimating Influence under Training Randomness
18. ICLR 2026FACT: a first-principles alternative to the Neural Feature Ansatz for how networks learn representations
19. ICLR 2026Faithfulness Under the Distribution: A New Look at Attribution Evaluation
20. ICLR 2026Feature compression is the root cause of adversarial fragility in neural networks
21. ICLR 2026FeatureBench: Benchmarking Agentic Coding for Complex Feature Development
22. ICLR 2026Flow-Disentangled Feature Importance
23. ICLR 2026GRADIEND: Feature Learning within Neural Networks Exemplified through Biases
24. ICLR 2026Hessian-Enhanced Token Attribution (HETA): Interpreting Autoregressive LLMs
25. ICLR 2026Influence Dynamics and Stagewise Data Attribution
26. ICLR 2026Intrinsic Explanation of Random Subspace Method for Enhanced Security Applications
27. ICLR 2026Learning to Weight Parameters for Training Data Attribution
28. ICLR 2026Let Features Decide Their Own Solvers: Hybrid Feature Caching for Diffusion Transformers
29. ICLR 2026Let OOD Feature Exploring Vast Predefined Classifiers
30. ICLR 2026Missingness Bias Calibration in Feature Attribution Explanations
31. ICLR 2026Mitigating the Curse of Detail: Scaling Arguments for Feature Learning and Sample Complexity
32. ICLR 2026Model Collapse Is Not a Bug but a Feature in Machine Unlearning for LLMs
33. ICLR 2026NIMO: a Nonlinear Interpretable MOdel
34. ICLR 2026Online time series prediction using feature adjustment
35. ICLR 2026Relational Feature Caching for Accelerating Diffusion Transformers
36. ICLR 2026Residual Feature Integration is Sufficient to Prevent Negative Transfer
37. ICLR 2026Scaling Attention via Feature Sparsity
38. ICLR 2026Segment-Level Attribution for Selective Learning of Long Reasoning Traces
39. ICLR 2026Splat Feature Solver
40. ICLR 2026Stable-LoRA: Stabilizing Feature Learning of Low-Rank Adaption
41. ICLR 2026Tackling the XAI Disagreement Problem with Adaptive Feature Grouping
42. ICLR 2026Taming Polysemanticity in LLMs: Theory-Grounded Feature Recovery via Sparse Autoencoders
43. ICLR 2026Temporal superposition and feature geometry of RNNs under memory demands
44. ICLR 2026Theoretical Analysis of Contrastive Learning under Imbalanced Data: From Training Dynamics to a Pruning Solution
45. ICLR 2026TIMESLIVER : SYMBOLIC-LINEAR DECOMPOSITION FOR EXPLAINABLE TIME SERIES CLASSIFICATION
46. ICLR 2026Transfer Learning in Infinite Width Feature Learning Networks
47. ICLR 2026Trapped by simplicity: When Transformers fail to learn from noisy features
48. ICLR 2026TreeGrad-Ranker: Feature Ranking via $O(L)$-Time Gradients for Decision Trees
49. ICLR 2026ULTRA-360: Unconstrained Dataset for Large-scale Temporal 3D Reconstruction across Altitudes and Omnidirectional Views
50. ICLR 2026Uncertainty as Feature Gaps: Epistemic Uncertainty Quantification of LLMs in Contextual Question-Answering

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。