Timezone: »

 
Poster
Beyond the Single Neuron Convex Barrier for Neural Network Certification
Gagandeep Singh · Rupanshu Ganvir · Markus Püschel · Martin Vechev

Thu Dec 12 10:45 AM -- 12:45 PM (PST) @ East Exhibition Hall B + C #121
in Deep Learning -- Adversarial Networks »

We propose a new parametric framework, called k-ReLU, for computing precise and scalable convex relaxations used to certify neural networks. The key idea is to approximate the output of multiple ReLUs in a layer jointly instead of separately. This joint relaxation captures dependencies between the inputs to different ReLUs in a layer and thus overcomes the convex barrier imposed by the single neuron triangle relaxation and its approximations. The framework is parametric in the number of k ReLUs it considers jointly and can be combined with existing verifiers in order to improve their precision. Our experimental results show that k-ReLU en- ables significantly more precise certification than existing state-of-the-art verifiers while maintaining scalability.

Keywords: Deep Learning -> Adversarial Networks · Algorithms -> Adversarial Learning

Author Information

Gagandeep Singh (ETH Zurich)
Rupanshu Ganvir (ETH Zurich)
Markus Püschel (ETH Zurich)
Martin Vechev (ETH Zurich, Switzerland)

More from the Same Authors