Transformers: A Security Perspective

BS Latibari, N Nazari, MA Chowdhury, KI Gubbi… - IEEE …, 2024 - ieeexplore.ieee.org
The Transformers architecture has recently emerged as a revolutionary paradigm in the field
of deep learning, particularly excelling in Natural Language Processing (NLP) and …

TPUXtract: An Exhaustive Hyperparameter Extraction Framework

A Kurian, A Dubey, F Yaman, A Aysu - IACR Transactions on …, 2025 - tches.iacr.org
Model stealing attacks on AI/ML devices undermine intellectual property rights, compromise
the competitive advantage of the original model developers, and potentially expose …

SoK: Model Reverse Engineering Threats for Neural Network Hardware

S Potluri, F Koushanfar - Cryptology ePrint Archive, 2024 - eprint.iacr.org
There has been significant progress over the past seven years in model reverse engineering
(RE) for neural network (NN) hardware. Although there has been systematization of …

Defense against ML-based Power Side-channel Attacks on DNN Accelerators with Adversarial Attacks

X Yan, CH Chang, T Zhang - arXiv preprint arXiv:2312.04035, 2023 - arxiv.org
Artificial Intelligence (AI) hardware accelerators have been widely adopted to enhance the
efficiency of deep learning applications. However, they also raise security concerns …

UniGuard: A Unified Hardware-oriented Threat Detector for FPGA-based AI Accelerators

X Yan, H Qiu, T Zhang - 2024 34th International Conference on …, 2024 - ieeexplore.ieee.org
The proliferation of AI technology gives rise to a variety of security threats, significantly
compromising the confidentiality and integrity of AI applications. Existing software-based …

A Unified Hardware-based Threat Detector for AI Accelerators

X Yan, H Qiu, T Zhang - arXiv preprint arXiv:2311.16684, 2023 - arxiv.org
The proliferation of AI technology gives rise to a variety of security threats, which significantly
compromise the confidentiality and integrity of AI models and applications. Existing software …

Theoretical models for physical security and their application

J Feldtkeller - 2024 - hss-opus.ub.ruhr-uni-bochum.de
Computertechnologien sind heute so sehr verwoben mit unserer Existenz, dass moderne
Türen sowohl physisch als auch metaphorisch durch einen Strom digitaler Daten geöffnet …