@inproceedings{HashemiKřetinskyRiederetal.2024, author = {Hashemi, Vahid and Křet{\´i}nsk{\´y}, Jan and Rieder, Sabine and Sch{\"o}n, Torsten and Vorhoff, Jan}, title = {Gaussian-Based and Outside-the-Box Runtime Monitoring Join Forces}, booktitle = {Runtime Verification: 24th International Conference, Proceedings}, editor = {{\´A}brah{\´a}m, Erika and Abbas, Houssam}, publisher = {Springer}, address = {Cham}, isbn = {978-3-031-74234-7}, doi = {https://doi.org/10.1007/978-3-031-74234-7_14}, pages = {218 -- 228}, year = {2024}, language = {en} } @unpublished{HashemiKřetinskyRiederetal.2024, author = {Hashemi, Vahid and Křet{\´i}nsk{\´y}, Jan and Rieder, Sabine and Sch{\"o}n, Torsten and Vorhoff, Jan}, title = {Gaussian-Based and Outside-the-Box Runtime Monitoring Join Forces}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2410.06051}, year = {2024}, abstract = {Since neural networks can make wrong predictions even with high confidence, monitoring their behavior at runtime is important, especially in safety-critical domains like autonomous driving. In this paper, we combine ideas from previous monitoring approaches based on observing the activation values of hidden neurons. In particular, we combine the Gaussian-based approach, which observes whether the current value of each monitored neuron is similar to typical values observed during training, and the Outside-the-Box monitor, which creates clusters of the acceptable activation values, and, thus, considers the correlations of the neurons' values. Our experiments evaluate the achieved improvement.}, language = {en} }