主动学习论文趋势 2023–2026

主要结论

• 主动学习论文从 2023 年的 58 篇降到 2026 年的 41 篇，热度在回落。
• NeurIPS 是发主动学习论文最多的会议，2023–2026 年共 67 篇，占四个会议总量的 41%。
• 交叉方向里涨得最快的是三维视觉：同时属于两个主题的论文从 2023 年的 0 篇增至 2026 年的 6 篇。
• 在「以数据为中心 AI」的 5 个主题里，它的 2026 年论文数排第 3；同组增长最快的合成数据生成是 1.3 倍，主动学习是 0.7 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 2026Active Corpus Selection for Training Subgraph Retrievers Using OOD Queries
• ICML 2026Active Regression for Single-Index Models with Unknown Link Functions
• ICLR 2026ActiveCQ: Active Estimation of Causal Quantities
• CVPR 2026Active Inference for Micro-Gesture Recognition EFE-Guided Temporal Sampling and Adaptive
• NeurIPS 2025Active Measurement: Efficient Estimation at Scale
• ICML 2025Active Learning of Deep Neural Networks via Gradient-Free Cutting Planes
• ICLR 2025Dual Process Learning: Controlling Use of In-Context vs. In-Weights Strategies with Weight Forgetting
• CVPR 2025Active Hyperspectral Imaging Using an Event Camera
• NeurIPS 2024On the Convergence of Loss and Uncertainty-based Active Learning Algorithms
• ICML 2024Active Statistical Inference
• ICLR 2024Improved Active Learning via Dependent Leverage Score Sampling
• CVPR 2024NARUTO: Neural Active Reconstruction from Uncertain Target Observations
• NeurIPS 2023A Competitive Algorithm for Agnostic Active Learning
• ICML 2023A Two-Stage Active Learning Algorithm for k-Nearest Neighbors
• ICLR 2023A Simple Yet Powerful Deep Active Learning With Snapshots Ensembles
• CVPR 2023Re-Thinking Federated Active Learning Based on Inter-Class Diversity
• NeurIPS 2026Active Learning as Nullspace Regulation: A Spectral Representation Perspective
• ICML 2026Active Timepoint Selection for Learning Measure-Valued Trajectories
• ICLR 2026TESSAR: Geometry-Aware Active Regression via Dynamic Voronoi Tessellation
• CVPR 2026EventHub Data Factory for Generalizable Event-Based Stereo Networks without Active

全部主动学习论文

共 58 篇，第 1 / 2 页

1. CVPR 2023A Light Weight Model for Active Speaker Detection
2. CVPR 2023Active Finetuning: Exploiting Annotation Budget in the Pretraining-Finetuning Paradigm
3. CVPR 2023AShapeFormer: Semantics-Guided Object-Level Active Shape Encoding for 3D Object Detection via Transformers
4. CVPR 2023Bi3D: Bi-Domain Active Learning for Cross-Domain 3D Object Detection
5. CVPR 2023Box-Level Active Detection
6. CVPR 2023Re-Thinking Federated Active Learning Based on Inter-Class Diversity
7. ICLR 2023A Simple Yet Powerful Deep Active Learning With Snapshots Ensembles
8. ICLR 2023Active Image Indexing
9. ICLR 2023Active Learning in Bayesian Neural Networks with Balanced Entropy Learning Principle
10. ICLR 2023BALTO: fast tensor program optimization with diversity-based active learning
11. ICLR 2023Leveraging Importance Weights in Subset Selection
12. ICLR 2023MCAL: Minimum Cost Human-Machine Active Labeling
13. ICLR 2023Robust Active Distillation
14. ICLR 2023Scalable Batch-Mode Deep Bayesian Active Learning via Equivalence Class Annealing
15. ICLR 2023Scalable Subset Sampling with Neural Conditional Poisson Networks
16. ICLR 2023Selective Annotation Makes Language Models Better Few-Shot Learners
17. ICML 2023A Two-Stage Active Learning Algorithm for k-Nearest Neighbors
18. ICML 2023Active Learning based Structural Inference
19. ICML 2023Active Policy Improvement from Multiple Black-box Oracles
20. ICML 2023Anchor Sampling for Federated Learning with Partial Client Participation
21. ICML 2023Gibbsian Polar Slice Sampling
22. ICML 2023LSDS++ : Dual Sampling for Accelerated k-means++
23. ICML 2023Margin-based sampling in high dimensions: When being active is less efficient than staying passive
24. ICML 2023NNSplitter: An Active Defense Solution for DNN Model via Automated Weight Obfuscation
25. ICML 2023On Sampling with Approximate Transport Maps
26. ICML 2023SAAL: Sharpness-Aware Active Learning
27. ICML 2023Streaming Active Learning with Deep Neural Networks
28. ICML 2023Towards Controlled Data Augmentations for Active Learning
29. ICML 2023Training-Free Neural Active Learning with Initialization-Robustness Guarantees
30. NeurIPS 2023A Competitive Algorithm for Agnostic Active Learning
31. NeurIPS 2023Active Bipartite Ranking
32. NeurIPS 2023Active Learning-Based Species Range Estimation
33. NeurIPS 2023Active Negative Loss Functions for Learning with Noisy Labels
34. NeurIPS 2023Active Vision Reinforcement Learning under Limited Visual Observability
35. NeurIPS 2023Actively Testing Your Model While It Learns: Realizing Label-Efficient Learning in Practice
36. NeurIPS 2023Adaptive Selective Sampling for Online Prediction with Experts
37. NeurIPS 2023Agnostic Multi-Group Active Learning
38. NeurIPS 2023Algorithm Selection for Deep Active Learning with Imbalanced Datasets
39. NeurIPS 2023Beyond Uniform Sampling: Offline Reinforcement Learning with Imbalanced Datasets
40. NeurIPS 2023Bucks for Buckets (B4B): Active Defenses Against Stealing Encoders
41. NeurIPS 2023Comparing Apples to Oranges: Learning Similarity Functions for Data Produced by Different Distributions
42. NeurIPS 2023Cross-modal Active Complementary Learning with Self-refining Correspondence
43. NeurIPS 2023DELTA: Diverse Client Sampling for Fasting Federated Learning
44. NeurIPS 2023Epidemic Learning: Boosting Decentralized Learning with Randomized Communication
45. NeurIPS 2023FIRAL: An Active Learning Algorithm for Multinomial Logistic Regression
46. NeurIPS 2023How Re-sampling Helps for Long-Tail Learning?
47. NeurIPS 2023How to Select Which Active Learning Strategy is Best Suited for Your Specific Problem and Budget
48. NeurIPS 2023Improving Language Plasticity via Pretraining with Active Forgetting
49. NeurIPS 2023Likelihood Ratio Confidence Sets for Sequential Decision Making
50. NeurIPS 2023Model-Free Active Exploration in Reinforcement Learning

常见问题

2023–2026 年四大会议一共有多少篇主动学习论文？

共 165 篇，按年份分别是 2023 年 58 篇、2024 年 35 篇、2025 年 31 篇、2026 年 41 篇。

NeurIPS 2026 有多少篇主动学习论文？

NeurIPS 2026 有 16 篇主动学习论文，上一届是 9 篇。

ICLR 2026 有多少篇主动学习论文？

ICLR 2026 有 6 篇主动学习论文，上一届是 7 篇。

ICML 2026 有多少篇主动学习论文？

ICML 2026 有 8 篇主动学习论文，上一届是 7 篇。

CVPR 2026 有多少篇主动学习论文？

CVPR 2026 有 11 篇主动学习论文，上一届是 8 篇。

哪个会议的主动学习论文最多？

NeurIPS，2023–2026 年共 67 篇。