A machine learning model is traditionally considered robust if its prediction remains (almost)
constant under input perturbations with small norm. However, real-world tasks like molecular …

Neural network controllers (NNCs) have shown great promise in autonomous and cyber-
physical systems. Despite the various verification approaches for neural networks, the safety …

Cooperative multi-agent reinforcement learning (c-MARL) is widely applied in safety-critical
scenarios, thus the analysis of robustness for c-MARL models is profoundly important …

Previous studies have shown that universal adversarial attacks can fool deep neural
networks over a large set of input images with a single human-invisible perturbation …

Deep neural networks (DNNs) are known to be vulnerable to adversarial geometric
transformation. This paper aims to verify the robustness of large-scale DNNs against the …

Reinforcement Learning (RL) has achieved remarkable success in safety-critical areas, but it
can be weakened by adversarial attacks. Recent studies have introduced``smoothed …

Deep neural networks have demonstrated remarkable performance in point cloud
classification. However previous works show they are vulnerable to adversarial …

The vulnerability of deep neural networks (DNNs) to input perturbations has posed a
significant challenge. Recent work on robustness verification of DNNs not only lacks …

Formal methods and machine learning are two research fields with drastically different
foundations and philosophies. Formal methods utilise mathematically rigorous techniques …

Verification plays an essential role in the formal analysis of safety-critical systems. Most
current verification methods have specific requirements when working on Deep Neural …