3 papers
cs.LG2026
Shifting-based Optimizable Linear Relaxations for General Activation Functions
Philipp Kern, László Antal, Erika Ábráham +1
The use of neural networks (NNs) is rapidly increasing, including in safety- and security-critical domains. To provide formal guarantees about NN behavior, many verification method…
cs.CR2026
Encrypted Neural Networks without Overflows
Philipp Kern, Lorenzo Rovida, Samuel Teuber +3
The popular Cheon-Kim-Kim-Song (CKKS) scheme enables efficient private inference in neural networks by evaluating them on encrypted data. Since CKKS only supports addition, multipl…
cs.LG2024
Revisiting Differential Verification: Equivalence Verification with Confidence
Samuel Teuber, Philipp Kern, Marvin Janzen +1
When validated neural networks (NNs) are pruned (and retrained) before deployment, it is desirable to prove that the new NN behaves equivalently to the (original) reference NN. To…