Topological Detection of Trojaned Neural Networks
This addresses security vulnerabilities in neural networks for applications requiring robust AI systems, though it is incremental as it builds on existing detection methods.
The paper tackles the problem of detecting Trojan attacks in deep neural networks by identifying structural deviations using topological tools, achieving better performance than standard baselines on multiple benchmarks.
Deep neural networks are known to have security issues. One particular threat is the Trojan attack. It occurs when the attackers stealthily manipulate the model's behavior through Trojaned training samples, which can later be exploited. Guided by basic neuroscientific principles we discover subtle -- yet critical -- structural deviation characterizing Trojaned models. In our analysis we use topological tools. They allow us to model high-order dependencies in the networks, robustly compare different networks, and localize structural abnormalities. One interesting observation is that Trojaned models develop short-cuts from input to output layers. Inspired by these observations, we devise a strategy for robust detection of Trojaned models. Compared to standard baselines it displays better performance on multiple benchmarks.