A particularly challenging problem in AI safety is providing guarantees ...
We propose a model-free reinforcement learning solution, namely the ASAP...
Models of actual causality leverage domain knowledge to generate convinc...
As machine learning models continue to achieve impressive performance ac...
Deep neural networks have repeatedly been shown to be non-robust to the
...
Uncertainty quantification and robustness to distribution shifts are
imp...
Generating accurate runtime safety estimates for autonomous systems is v...
Although organizations are continuously making concerted efforts to hard...
Machine learning models are prone to making incorrect predictions on inp...
Adversarial training (AT) and its variants have spearheaded progress in
...
Anomaly detection is essential for preventing hazardous outcomes for
saf...
This paper tackles the problem of making complex resource-constrained
cy...
Machine learning methods such as deep neural networks (DNNs), despite th...
Autonomous systems with machine learning-based perception can exhibit
un...
Closed-loop verification of cyber-physical systems with neural network
c...
Deep neural networks (DNNs) are known to produce incorrect predictions w...
Improving adversarial robustness of neural networks remains a major
chal...
Deep neural networks (DNNs) are known to produce incorrect predictions w...
Industrial cyber-physical systems are hybrid systems with strict safety
...
In this position paper we discuss three main shortcomings of existing
ap...