mlsanityData to 7 Oct 2026

HAMLOCK: HArdware-Model LOgically Combined attacK

arXiv 2510.19145 ↗First posted 22 Oct 2025cs.CR cs.LGcited by 0

On the map: LLM safety and security → Backdoor and data poisoning attacks

Abstract (arXiv)

The growing use of third-party hardware accelerators (e.g., FPGAs, ASICs) for deep neural networks (DNNs) introduces new security vulnerabilities. Conventional model-level backdoor attacks, which only poison a model's weights to misclassify inputs with a specific trigger, are often detectable because the entire attack logic is embedded within the model (i.e., software), creating a traceable layer-by-layer activation path. This paper introduces the HArdware-Model Logically Combined Attack (HAMLOCK), a far stealthier threat that distributes the attack logic across the hardware-software boundary. The software (model) is now only minimally altered by tuning the activations of few neurons to produce uniquely high activation values when a trigger is present. A malicious hardware Trojan detects those unique activations by monitoring the corresponding neurons' most significant bit or the 8-bit exponents and triggers another hardware Trojan to directly manipulate the final output logits for misclassification. This decoupled design is highly stealthy, as the model itself contains no complete backdoor activation path as in conventional attacks and hence, appears fully benign. Empirically, across benchmarks like MNIST, CIFAR10, GTSRB, and ImageNet, HAMLOCK achieves a near-perfect attack success rate with a negligible clean accuracy drop. More importantly, HAMLOCK circumvents the state-of-the-art model-level defenses without any adaptive optimization. The hardware Trojan is also undetectable, incurring area and power overheads as low as 0.01%, which is easily masked by process and environmental noise. Our findings expose a critical vulnerability at the hardware-software interface, demanding new cross-layer defenses against this emerging threat.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Backdoor and data poisoning attacks, one of 12 directions in LLM safety and security. Builds on 39 of its 78 references that are in the map. First posted on arXiv 22 Oct 2025.

Builds on

39 of 78 references.

Closest ancestors · 1

Same direction, newest first

  1. 2023Setting the Trap: Capturing and Defeating Backdoors in Pretrained Language Models through Honeypots28 Oct

Other references · 38

  1. 2017Attention Is All You Needcited by 59,932
  2. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  3. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  4. 2015Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Codingcited by 2,291
  5. 2017Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoningcited by 1,012
  6. 2018Fine-Pruning: Defending Against Backdooring Attacks on Deep Neural Networkscited by 550
  7. 2018Spectral Signatures in Backdoor Attackscited by 457
  8. 2019STRIP: A Defence Against Trojan Attacks on Deep Neural Networkscited by 437
  9. 2019A Survey of Deep Learning Techniques for Autonomous Drivingcited by 401
  10. 2019Hidden Trigger Backdoor Attackscited by 359
  11. 2021Neural Attention Distillation: Erasing Backdoor Triggers from Deep Neural Networkscited by 313
  12. 2021Anti-Backdoor Learning: Training Clean Models on Poisoned Datacited by 258
26 more
  1. 2019Detecting AI Trojans Using Meta Neural Analysiscited by 192
  2. 2021Adversarial Unlearning of Backdoors via Implicit Hypergradientcited by 160
  3. 2022Backdoor Defense via Decoupling the Training Processcited by 149
  4. 2021SPECTRE: Defending Against Backdoor Attacks Using Robust Statisticscited by 123
  5. 2023SCALE-UP: An Efficient Black-box Input-level Backdoor Detection via Analyzing Scaled Prediction Consistencycited by 93
  6. 2020An Embarrassingly Simple Approach for Trojan Attack in Deep Neural Networkscited by 92
  7. 2019Design and Evaluation of a Multi-Domain Trojan Detection Method on Deep Neural Networkscited by 66
  8. 2022The "Beatrix'' Resurrections: Robust Backdoor Detection via Gram Matricescited by 62
  9. 2021Handcrafted Backdoors in Deep Neural Networkscited by 57
  10. 2019Terminal Brain Damage: Exposing the Graceless Degradation in Deep Neural Networks Under Hardware Fault Attackscited by 55
  11. 2022MM-BD: Post-Training Detection of Backdoor Attacks with Arbitrary Backdoor Pattern Types Using a Maximum Margin Statisticcited by 53
  12. 2023Detecting Backdoors During the Inference Stage Based on Corruption Robustness Consistencycited by 45
  13. 2022Training with More Confidence: Mitigating Injected and Natural Backdoors During Trainingcited by 38
  14. 2024IBD-PSC: Input-level Backdoor Detection via Parameter-oriented Scaling Consistencycited by 38
  15. 2022Fine-Tuning Is All You Need to Mitigate Backdoor Attackscited by 35
  16. 2022Architectural Backdoors in Neural Networkscited by 24
  17. 2023Robust Backdoor Detection for Deep Learning via Topological Evolution Dynamicscited by 24
  18. 2021Stealthy Backdoors as Compression Artifactscited by 17
  19. 2023UMD: Unsupervised Model Detection for X2X Backdoor Attackscited by 17
  20. 2023BEAGLE: Forensics of Deep Learning Backdoor Attack for Better Defensecited by 15
  21. 2023Defending Against Backdoor Attacks by Layer-wise Feature Analysiscited by 15
  22. 2024Nearest is Not Dearest: Towards Practical Defense against Quantization-conditioned Backdoor Attackscited by 14
  23. 2018Hardware Trojan Attacks on Neural Networkscited by 11
  24. 2018Hu-Fu: Hardware and Software Collaborative Attack Framework against Neural Networkscited by 8
  25. 2024Data Free Backdoor Attackscited by 5
  26. 2025Democratic Training Against Universal Adversarial Perturbationscited by 1

Built on by

Not cited yet.