后门攻击论文趋势 2023–2026

主要结论

• 后门攻击论文从 2023 年的 73 篇增长到 2026 年的 111 篇，三年间扩大到约 1.5 倍。
• NeurIPS 是发后门攻击论文最多的会议，2023–2026 年共 124 篇，占四个会议总量的 33%。
• 交叉方向里涨得最快的是对齐与安全：同时属于两个主题的论文从 2023 年的 0 篇增至 2026 年的 16 篇。
• 在「安全与可信」的 5 个主题里，它的 2026 年论文数排第 2；同组增长最快的水印是 9.3 倍，后门攻击是 1.5 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 2026A Theoretical Analysis of Backdoor Learning as Simplicity-Biased Optimization Dynamics
• ICML 2026Coupled Trigger Optimization and Vulnerable Parameter Alignment for Persistent Backdoor Attacks on Federated Learning
• ICLR 2026Defending against Backdoor Attacks via Module Switching
• CVPR 2026Eliminate Distance Differences Induced by Backdoor Attacks Layer-Selective Training and
• NeurIPS 2025A Set of Generalized Components to Achieve Effective Poison-only Clean-label Backdoor Attacks with Collaborative Sample Selection and Triggers
• ICML 2025A Closer Look at Backdoor Attacks on CLIP
• ICLR 2025Activation Gradient based Poisoned Sample Detection Against Backdoor Attacks
• CVPR 2025BadToken: Token-level Backdoor Attacks to Multi-modal Large Language Models
• NeurIPS 2024Reproducibility Study on Adversarial Attacks Against Robust Transformer Trackers
• ICML 2024Better Safe than Sorry: Pre-training CLIP against Targeted Data Poisoning and Backdoor Attacks
• ICLR 2024Adversarial Feature Map Pruning for Backdoor
• CVPR 2024Data Poisoning based Backdoor Attacks to Contrastive Learning
• NeurIPS 2023A Unified Detection Framework for Inference-Stage Backdoor Defenses
• ICML 2023Data Poisoning Attacks Against Multimodal Encoders
• ICLR 2023Adversarial Attacks on Adversarial Bandits
• CVPR 2023Architectural Backdoors in Neural Networks
• NeurIPS 2026AI Models Can Provably Hide Arbitrary Capabilities
• ICML 2026DF-LoGiT: Data-Free Logic-Gated Backdoor Attacks in Vision Transformers
• ICLR 2026Don't Shift the Trigger: Robust Gradient Ascent for Backdoor Unlearning
• CVPR 2026Logit-Margin Repulsion for Backdoor Defense

全部后门攻击论文

共 73 篇，第 1 / 2 页

1. CVPR 2023Architectural Backdoors in Neural Networks
2. CVPR 2023Backdoor Attacks Against Deep Image Compression via Adaptive Frequency Trigger
3. CVPR 2023Backdoor Cleansing With Unlabeled Data
4. CVPR 2023Backdoor Defense via Adaptively Splitting Poisoned Dataset
5. CVPR 2023Backdoor Defense via Deconfounded Representation Learning
6. CVPR 2023Can't Steal? Cont-Steal! Contrastive Stealing Attacks Against Image Encoders
7. CVPR 2023Color Backdoor: A Robust Poisoning Attack in Color Space
8. CVPR 2023Defending Against Patch-Based Backdoor Attacks on Self-Supervised Learning
9. CVPR 2023Detecting Backdoors During the Inference Stage Based on Corruption Robustness Consistency
10. CVPR 2023Detecting Backdoors in Pre-Trained Encoders
11. CVPR 2023Dynamic Generative Targeted Attacks With Pattern Injection
12. CVPR 2023Efficient Loss Function by Minimizing the Detrimental Effect of Floating-Point Errors on Gradient-Based Attacks
13. CVPR 2023Enhancing the Self-Universality for Transferable Targeted Attacks
14. CVPR 2023Ensemble-Based Blackbox Attacks on Dense Prediction
15. CVPR 2023How to Backdoor Diffusion Models?
16. CVPR 2023MEDIC: Remove Model Backdoors via Importance Driven Cloning
17. CVPR 2023Minimizing Maximum Model Discrepancy for Transferable Black-Box Targeted Attacks
18. CVPR 2023Progressive Backdoor Erasing via Connecting Backdoor and Adversarial Attacks
19. CVPR 2023Re-Thinking Model Inversion Attacks Against Deep Neural Networks
20. CVPR 2023Reinforcement Learning-Based Black-Box Model Inversion Attacks
21. CVPR 2023Sibling-Attack: Rethinking Transferable Adversarial Attacks Against Face Recognition
22. CVPR 2023Single Image Backdoor Inversion via Robust Smoothed Classifiers
23. CVPR 2023The Best Defense Is a Good Offense: Adversarial Augmentation Against Adversarial Attacks
24. CVPR 2023The Dark Side of Dynamic Routing Neural Networks: Towards Efficiency Backdoor Injection
25. CVPR 2023TrojDiff: Trojan Attacks on Diffusion Models With Diverse Targets
26. CVPR 2023You Are Catching My Attention: Are Vision Transformers Bad Learners Under Backdoor Attacks?
27. ICLR 2023Adversarial Attacks on Adversarial Bandits
28. ICLR 2023Certified Defences Against Adversarial Patch Attacks on Semantic Segmentation
29. ICLR 2023Clean-image Backdoor: Attacking Multi-label Models with Poisoned Labels Only
30. ICLR 2023Distilling Cognitive Backdoor Patterns within an Image
31. ICLR 2023Effective passive membership inference attacks in federated learning against overparameterized models
32. ICLR 2023Few-shot Backdoor Attacks via Neural Tangent Kernels
33. ICLR 2023FLIP: A Provable Defense Framework for Backdoor Mitigation in Federated Learning
34. ICLR 2023Incompatibility Clustering as a Defense Against Backdoor Poisoning Attacks
35. ICLR 2023Indiscriminate Poisoning Attacks on Unsupervised Contrastive Learning
36. ICLR 2023Revisiting the Assumption of Latent Separability for Backdoor Defenses
37. ICLR 2023SCALE-UP: An Efficient Black-box Input-level Backdoor Detection via Analyzing Scaled Prediction Consistency
38. ICLR 2023The Dark Side of AutoML: Towards Architectural Backdoor Search
39. ICLR 2023Towards Robustness Certification Against Universal Perturbations
40. ICLR 2023TrojText: Test-time Invisible Textual Trojan Insertion
41. ICLR 2023UNICORN: A Unified Backdoor Trigger Inversion Framework
42. ICML 2023Are Diffusion Models Vulnerable to Membership Inference Attacks?
43. ICML 2023Data Poisoning Attacks Against Multimodal Encoders
44. ICML 2023Exploring the Limits of Model-Targeted Indiscriminate Data Poisoning Attacks
45. ICML 2023Graph Contrastive Backdoor Attacks
46. ICML 2023MultiRobustBench: Benchmarking Robustness Against Multiple Attacks
47. ICML 2023Preprocessors Matter! Realistic Decision-Based Attacks on Machine Learning Systems
48. ICML 2023Reconstructive Neuron Pruning for Backdoor Defense
49. ICML 2023Rethinking Backdoor Attacks
50. ICML 2023Run-off Election: Improved Provable Defense against Data Poisoning Attacks

常见问题

2023–2026 年四大会议一共有多少篇后门攻击论文？

共 381 篇，按年份分别是 2023 年 73 篇、2024 年 105 篇、2025 年 92 篇、2026 年 111 篇。

NeurIPS 2026 有多少篇后门攻击论文？

NeurIPS 2026 有 36 篇后门攻击论文，上一届是 31 篇。

ICLR 2026 有多少篇后门攻击论文？

ICLR 2026 有 20 篇后门攻击论文，上一届是 25 篇。

ICML 2026 有多少篇后门攻击论文？

ICML 2026 有 36 篇后门攻击论文，上一届是 17 篇。

CVPR 2026 有多少篇后门攻击论文？

CVPR 2026 有 19 篇后门攻击论文，上一届是 19 篇。

哪个会议的后门攻击论文最多？

NeurIPS，2023–2026 年共 124 篇。