TY - GEN
T1 - Symbolic Gaussian Smoothing
AU - Mata, Andrew
AU - Arjomandbigdeli, Ali
AU - Bak, Stanley
N1 - Publisher Copyright:
© 2025 Copyright held by the owner/author(s).
PY - 2025/5/7
Y1 - 2025/5/7
N2 - This work introduces a deterministic variant of randomized smoothing for certifying neural network robustness. Instead of relying on multiple stochastic forward passes with Gaussian noise, the proposed method employs symbolic moment propagation to analytically compute the exact mean and covariance of the output distribution in a single pass. Evaluations on MNIST and Fashion MNIST show that the method yields comparable classification accuracy while achieving, on average, 10% larger certified robustness radii. This deterministic framework offers exact, non-probabilistic guarantees that are particularly attractive for safety-critical applications.
AB - This work introduces a deterministic variant of randomized smoothing for certifying neural network robustness. Instead of relying on multiple stochastic forward passes with Gaussian noise, the proposed method employs symbolic moment propagation to analytically compute the exact mean and covariance of the output distribution in a single pass. Evaluations on MNIST and Fashion MNIST show that the method yields comparable classification accuracy while achieving, on average, 10% larger certified robustness radii. This deterministic framework offers exact, non-probabilistic guarantees that are particularly attractive for safety-critical applications.
KW - Certified Defense
KW - Neural Network Robustness
UR - https://www.scopus.com/pages/publications/105007286461
U2 - 10.1145/3716550.3725169
DO - 10.1145/3716550.3725169
M3 - Conference contribution
AN - SCOPUS:105007286461
T3 - Proceedings of the ACM/IEEE 16th International Conference on Cyber-Physical Systems, ICCPS 2025, held as part of the CPS-IoT Week 2025
BT - Proceedings of the ACM/IEEE 16th International Conference on Cyber-Physical Systems, ICCPS 2025, held as part of the CPS-IoT Week 2025
PB - Association for Computing Machinery, Inc
T2 - 16th Annual ACM/IEEE International Conference on Cyber-Physical Systems, ICCPS 2025, held as part of the CPS-IoT Week 2025
Y2 - 6 May 2025 through 9 May 2025
ER -