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 695 篇，第 8 / 14 页

351. ICML 2025Towards an Explainable Comparison and Alignment of Feature Embeddings
352. ICML 2025Weak-to-Strong Generalization Even in Random Feature Networks, Provably
353. NeurIPS 2025A is for Absorption: Studying Feature Splitting and Absorption in Sparse Autoencoders
354. NeurIPS 2025Accelerating Feature Conformal Prediction via Taylor Approximation
355. NeurIPS 2025Alternating Gradient Flows: A Theory of Feature Learning in Two-layer Neural Networks
356. NeurIPS 2025Approximating Shapley Explanations in Reinforcement Learning
357. NeurIPS 2025Attribution-Driven Adaptive Token Pruning for Transformers
358. NeurIPS 2025Better Training Data Attribution via Better Inverse Hessian-Vector Products
359. NeurIPS 2025BevSplat: Resolving Height Ambiguity via Feature-Based Gaussian Primitives for Weakly-Supervised Cross-View Localization
360. NeurIPS 2025Beyond Last-Click: An Optimal Mechanism for Ad Attribution
361. NeurIPS 2025Capturing Polysemanticity with PRISM: A Multi-Concept Feature Description Framework
362. NeurIPS 2025Closed-Form Training Dynamics Reveal Learned Features and Linear Structure in Word2Vec-like Models
363. NeurIPS 2025Contrastive Learning with Data Misalignment: Feature Purity, Training Dynamics and Theoretical Generalization Guarantees
364. NeurIPS 2025DATE-LM: Benchmarking Data Attribution Evaluation for Large Language Models
365. NeurIPS 2025Degrees of Freedom for Linear Attention: Distilling Softmax Attention with Optimal Feature Efficiency
366. NeurIPS 2025Demystifying Spectral Feature Learning for Instrumental Variable Regression
367. NeurIPS 2025DePass: Unified Feature Attributing by Simple Decomposed Forward Pass
368. NeurIPS 2025Distributional Training Data Attribution: What do Influence Functions Sample?
369. NeurIPS 2025Distributionally Robust Feature Selection
370. NeurIPS 2025Dynamical Decoupling of Generalization and Overfitting in Large Two-Layer Networks
371. NeurIPS 2025Efficiently Verifiable Proofs of Data Attribution
372. NeurIPS 2025Enhancing Training Data Attribution with Representational Optimization
373. NeurIPS 2025Explaining Similarity in Vision-Language Encoders with Weighted Banzhaf Interactions
374. NeurIPS 2025Exploration via Feature Perturbation in Contextual Bandits
375. NeurIPS 2025Fast Data Attribution for Text-to-Image Models
376. NeurIPS 2025Feature Distillation is the Better Choice for Model-Heterogeneous Federated Learning
377. NeurIPS 2025Feature-aware Modulation for Learning from Temporal Tabular Data
378. NeurIPS 2025Final-Model-Only Data Attribution with a Unifying View of Gradient-Based Methods
379. NeurIPS 2025From Linear to Nonlinear: Provable Weak-to-Strong Generalization through Feature Learning
380. NeurIPS 2025GraSS: Scalable Data Attribution with Gradient Sparsification and Sparse Projection
381. NeurIPS 2025H-SPLID: HSIC-based Saliency Preserving Latent Information Decomposition
382. NeurIPS 2025Harnessing Feature Resonance under Arbitrary Target Alignment for Out-of-Distribution Node Detection
383. NeurIPS 2025How Classifier Features Transfer to Downstream: An Asymptotic Analysis in a Two-Layer Model
384. NeurIPS 2025ImageNet-trained CNNs are not biased towards texture: Revisiting feature reliance through controlled suppression
385. NeurIPS 2025Inverse Methods for Missing Data Imputation
386. NeurIPS 2025JAFAR: Jack up Any Feature at Any Resolution
387. NeurIPS 2025Neural Collapse in Cumulative Link Models for Ordinal Regression: An Analysis with Unconstrained Feature Model
388. NeurIPS 2025On the Mechanisms of Weak-to-Strong Generalization: A Theoretical Perspective
389. NeurIPS 2025On the Surprising Effectiveness of Large Learning Rates under Standard Width Scaling
390. NeurIPS 2025OrdShap: Feature Position Importance for Sequential Black-Box Models
391. NeurIPS 2025Plug-and-play Feature Causality Decomposition for Multimodal Representation Learning
392. NeurIPS 2025Practical do-Shapley Explanations with Estimand-Agnostic Causal Inference
393. NeurIPS 2025Proxy-SPEX: Sample-Efficient Interpretability via Sparse Feature Interactions in LLMs
394. NeurIPS 2025Regional Explanations: Bridging Local and Global Variable Importance
395. NeurIPS 2025Rescaled Influence Functions: Accurate Data Attribution in High Dimension
396. NeurIPS 2025Revising and Falsifying Sparse Autoencoder Feature Explanations
397. NeurIPS 2025Revisiting LRP: Positional Attribution as the Missing Ingredient for Transformer Explainability
398. NeurIPS 2025Scalable Feature Learning on Huge Knowledge Graphs for Downstream Machine Learning
399. NeurIPS 2025Sculpting Features from Noise: Reward-Guided Hierarchical Diffusion for Task-Optimal Feature Transformation
400. NeurIPS 2025SHAP values via sparse Fourier representation

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。