Paper compares explainability methods for hardware Trojan detection

A new paper published on arXiv details a systematic comparison of explainability methods for detecting hardware Trojans in integrated circuits. The research evaluates three categories of techniques: domain-aware property-based analysis, model-agnostic case-based reasoning, and model-agnostic feature attribution methods like LIME and SHAP. The study aims to determine which methods provide the most actionable insights for hardware engineers in identifying malicious circuits, which are difficult to fix once manufactured. The comparison was conducted using the Trust-Hub benchmark dataset. AI

IMPACT Provides insights into improving the security of integrated circuits through better AI-driven detection of malicious hardware components.

RANK_REASON The item is a research paper published on arXiv comparing explainability methods for a specific technical problem. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• alphaXiv
• arXiv
• CatalyzeX Code Finder for Papers
• Connected Papers
• DagsHub
• Gotit.pub
• Hardware Trojans Detection Based on Electromagnetic Emission Signals Analysis
• Hugging Face
• Influence Flower
• Litmaps
• Paul F. Whitten
• ScienceCast
• scite Smart Citations
• Shap
• Trust-Hub

• paper
• safety

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →