Search

Pushmeet Kohli

The Autoencoding Variational Autoencoder
Enabling certification of verification-agnostic networks via memory-efficient semidefinite programming
Toward Evaluating Robustness of Deep Reinforcement Learning with Continuous Control
Towards Verified Robustness under Text Deletion Interventions
A FRAMEWORK FOR ROBUSTNESS CERTIFICATION OF SMOOTHED CLASSIFIERS USING F-DIVERGENCES
Adversarially Robust Representations with Smooth Encoders
Adversarial Robustness through Local Linearization
Efficient Neural Network Verification with Exactness Characterization
Knowing When to Stop: Evaluation and Verification of Conformity to Output-size Specs
Rigorous Agent Evaluation: An Adversarial Approach to Uncover Catastrophic Failures
Verification of Non-Linear Specifications for Neural Networks
Achieving Verified Robustness to Symbol Substitutions via Interval Bound Propagation
Scalable Verified Training for Provably Robust Image Classification
Verification of deep probabilistic models
A dual approach to scalable verification of deep networks

Published with Wowchemy — the free, open source website builder that empowers creators.