mlsanityData to 7 Oct 2026

UNIT: Backdoor Mitigation via Automated Neural Distribution Tightening

arXiv 2407.11372 ↗First posted 16 Jul 2024cs.CR cs.CVcited by 2

Not on the map. Why?

Abstract (arXiv)

Deep neural networks (DNNs) have demonstrated effectiveness in various fields. However, DNNs are vulnerable to backdoor attacks, which inject a unique pattern, called trigger, into the input to cause misclassification to an attack-chosen target label. While existing works have proposed various methods to mitigate backdoor effects in poisoned models, they tend to be less effective against recent advanced attacks. In this paper, we introduce a novel post-training defense technique UNIT that can effectively eliminate backdoor effects for a variety of attacks. In specific, UNIT approximates a unique and tight activation distribution for each neuron in the model. It then proactively dispels substantially large activation values that exceed the approximated boundaries. Our experimental results demonstrate that UNIT outperforms 7 popular defense methods against 14 existing backdoor attacks, including 2 advanced attacks, using only 5% of clean training data. UNIT is also cost efficient. The code is accessible at https://github.com/Megum1/UNIT.

Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 57 of its 82 references that are in the map. First posted on arXiv 16 Jul 2024.

Builds on

57 of 82 references.

References · 57

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  3. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  4. 2015Distilling the Knowledge in a Neural Networkcited by 10,076
  5. 2016Densely Connected Convolutional Networkscited by 6,691
  6. 2017A Unified Approach to Interpreting Model Predictionscited by 6,488
  7. 2016Overcoming catastrophic forgetting in neural networkscited by 4,718
  8. 2017MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applicationscited by 3,827
  9. 2016Wide Residual Networkscited by 3,057
  10. 2015Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs)cited by 1,061
  11. 2017BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chaincited by 1,052
  12. 2017Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoningcited by 1,012
38 more
  1. 2018How To Backdoor Federated Learningcited by 745
  2. 2017Self-Normalizing Neural Networkscited by 591
  3. 2018Fine-Pruning: Defending Against Backdooring Attacks on Deep Neural Networkscited by 550
  4. 2018Spectral Signatures in Backdoor Attackscited by 457
  5. 2019STRIP: A Defence Against Trojan Attacks on Deep Neural Networkscited by 437
  6. 2018Detecting Backdoor Attacks on Deep Neural Networks by Activation Clusteringcited by 429
  7. 2019Hidden Trigger Backdoor Attackscited by 359
  8. 2020Reflection Backdoor: A Natural Backdoor Attack on Deep Neural Networkscited by 341
  9. 2020Invisible Backdoor Attack with Sample-Specific Triggerscited by 338
  10. 2021Neural Attention Distillation: Erasing Backdoor Triggers from Deep Neural Networkscited by 313
  11. 2020Input-Aware Dynamic Backdoor Attackcited by 304
  12. 2021Anti-Backdoor Learning: Training Clean Models on Poisoned Datacited by 258
  13. 2019A new Backdoor Attack in CNNs by training set corruption without label poisoningcited by 237
  14. 2021Adversarial Neuron Pruning Purifies Backdoored Deep Modelscited by 230
  15. 2021Hidden Killer: Invisible Textual Backdoor Attacks with Syntactic Triggercited by 203
  16. 2019Detecting AI Trojans Using Meta Neural Analysiscited by 192
  17. 2021Adversarial Unlearning of Backdoors via Implicit Hypergradientcited by 160
  18. 2021Activation Functions in Deep Learning: A Comprehensive Survey and Benchmarkcited by 153
  19. 2022Backdoor Defense via Decoupling the Training Processcited by 149
  20. 2020Dynamic Backdoor Attacks Against Machine Learning Modelscited by 147
  21. 2021SPECTRE: Defending Against Backdoor Attacks Using Robust Statisticscited by 123
  22. 2019TABOR: A Highly Accurate Approach to Inspecting and Restoring Trojan Backdoors in AI Systemscited by 120
  23. 2021BadEncoder: Backdoor Attacks to Pre-trained Encoders in Self-Supervised Learningcited by 120
  24. 2022Narcissus: A Practical Clean-Label Backdoor Attack with Limited Informationcited by 102
  25. 2022How to Backdoor Diffusion Models?cited by 102
  26. 2021Sleeper Agent: Scalable Hidden Trigger Backdoors for Neural Networks Trained from Scratchcited by 98
  27. 2022Neurotoxin: Durable Backdoors in Federated Learningcited by 97
  28. 2021Backdoor Scanning for Deep Neural Networks through K-Arm Optimizationcited by 96
  29. 2020Practical Detection of Trojan Neural Networks: Data-Limited and Data-Free Casescited by 93
  30. 2020Deep Feature Space Trojan Attack of Neural Networks by Controlled Detoxificationcited by 93
  31. 2022Data-free Backdoor Removal based on Channel Lipschitznesscited by 77
  32. 2023Enhancing Fine-Tuning Based Backdoor Defense with Sharpness-Aware Minimizationcited by 60
  33. 2023Reconstructive Neuron Pruning for Backdoor Defensecited by 56
  34. 2023Detecting Backdoors in Pre-trained Encoderscited by 53
  35. 2022Constrained Optimization with Dynamic Bound-scaling for Effective NLPBackdoor Defensecited by 44
  36. 2022Training with More Confidence: Mitigating Injected and Natural Backdoors During Trainingcited by 38
  37. 2022FLIP: A Provable Defense Framework for Backdoor Mitigation in Federated Learningcited by 37
  38. 2022Fine-Tuning Is All You Need to Mitigate Backdoor Attackscited by 35

50 most cited of 57 All 57 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 1; 2026: 1.

2 papers.

  1. 2025Defending against Backdoor Attacks via Module Switchingcited by 0
  2. 2026Kill it with FIRE: On Leveraging Latent Space Directions for Runtime Backdoor Mitigation in Deep Neural Networkscited by 0