The answer lies within: Detecting Trojans from DNNs' inherent characteristics

doi:10.1016/j.neunet.2026.108573

CSpace

	The answer lies within: Detecting Trojans from DNNs' inherent characteristics
	Liu, Xuchao; Cao, Qi; Zhang, Kaike; Su, Du; Shen, Huawei
	2026-06-01
发表期刊	NEURAL NETWORKS
ISSN	0893-6080
卷号	198 页码:11
摘要	Deep neural networks (DNNs) are vulnerable to Trojan attacks, where adversaries implant Trojans that cause DNNs to misbehave when encountering specific triggers. Detecting Trojans in DNNs is crucial to mitigate potential safety risks. Traditional methods typically employ trigger reversion techniques, which utilize benign samples to reconstruct potential triggers through iterative optimization. However, their practical applicability is limited by reliance on benign samples and the exceedingly time-intensive optimization. In this paper, we investigate more general yet challenging setting, the benign sample-free scenario, where detection relies solely on DNN itself. We propose a novel approach for detecting Trojans from DNNs' inherent characteristics (DTIC), which exploits the distinguishable features of Trojaned models. DTIC depicts the characteristics of various DNNs via a unified representation space derived from both views of model structures and parameters, enabling adaptability across diverse DNNs. It requires just one direct inference to assess the presence of Trojans, ensuring high efficiency. We further enhance the performance of Trojan detection, using augmentations based on random perturbations and the lottery hypothesis. Extensive experiments conducted on IARPA TrajAI1, a widely adopted benchmark, demonstrate the superior effectiveness, efficiency, and generalizability of DTIC.
关键词	Trojan detection Sample-free DNNs' inherent characteristics
DOI	10.1016/j.neunet.2026.108573
收录类别	SCI
语种	英语
WOS研究方向	Computer Science ; Neurosciences & Neurology
WOS类目	Computer Science, Artificial Intelligence ; Neurosciences
WOS记录号	WOS:001668510200001
出版者	PERGAMON-ELSEVIER SCIENCE LTD
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/42844
专题	中国科学院计算技术研究所
通讯作者	Cao, Qi
作者单位	Chinese Acad Sci, Inst Comp Technol, Beijing 100190, Peoples R China
推荐引用方式 GB/T 7714	Liu, Xuchao,Cao, Qi,Zhang, Kaike,et al. The answer lies within: Detecting Trojans from DNNs' inherent characteristics[J]. NEURAL NETWORKS,2026,198:11.
APA	Liu, Xuchao,Cao, Qi,Zhang, Kaike,Su, Du,&Shen, Huawei.(2026).The answer lies within: Detecting Trojans from DNNs' inherent characteristics.NEURAL NETWORKS,198,11.
MLA	Liu, Xuchao,et al."The answer lies within: Detecting Trojans from DNNs' inherent characteristics".NEURAL NETWORKS 198(2026):11.