A dual-penalty training loss that suppresses trigger-feature gradients hides tabular backdoors from Integrated Gradients, DeepSHAP, and CAD-Detect while maintaining high attack success.
ACM Computing Surveys , volume=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Crushing the Evidence: A Dual-Penalty Evasion Framework for Fooling White-Box Explainable AI Auditors
A dual-penalty training loss that suppresses trigger-feature gradients hides tabular backdoors from Integrated Gradients, DeepSHAP, and CAD-Detect while maintaining high attack success.