AI Safety

Handling Black Swan Events in Deep Learning with Diversely Extrapolated Neural Networks IJCAI 2020 Hidden Risks of Machine Learning Applied to Healthcare: Unintended Feedback Loops Between Models and Future Data Causing Model Degradation MLHC 2020 Hierarchical Verification for Adversarial Robustness ICML 2020 Higher-Order Certification For Randomized Smoothing NeurIPS 2020 How to 0wn the NAS in Your Spare Time ICLR 2020 Imitation Attacks and Defenses for Black-box Machine Translation Systems EMNLP 2020 Improving Adversarial Robustness Requires Revisiting Misclassified Examples ICLR 2020 Improving Robustness via Risk Averse Distributional Reinforcement Learning L4DC 2020 Incorporating Failure Events in Agents’ Decision Making to Improve User Satisfaction IJCAI 2020 Investigating Robustness of Adversarial Samples Detection for Automatic Speaker Verification Interspeech 2020 Lagrangian Decomposition for Neural Network Verification UAI 2020 Learn2Perturb: An End-to-End Feature Perturbation Learning to Improve Adversarial Robustness CVPR 2020 Learning Black-Box Attackers with Transferable Priors and Query Feedback NeurIPS 2020 Learning for Safety-Critical Control with Control Barrier Functions L4DC 2020 Learning from Interventions Using Hierarchical Policies for Safe Learning AAAI 2020 Learning Human Objectives by Evaluating Hypothetical Behavior ICML 2020 Lipschitz constant estimation of Neural Networks via sparse polynomial optimization ICLR 2020 LSTM Neural Networks: Input to State Stability and Probabilistic Safety Verification L4DC 2020 MACER: Attack-free and Scalable Robust Training via Maximizing Certified Radius ICLR 2020 MetaPoison: Practical General-purpose Clean-label Data Poisoning NeurIPS 2020 Mixup Inference: Better Exploiting Mixup to Defend Adversarial Attacks ICLR 2020 ML-LOO: Detecting Adversarial Examples with Feature Attribution AAAI 2020 Network Error Logging: Client-side measurement of end-to-end web service reliability NSDI 2020 Neural Network Branching for Neural Network Verification ICLR 2020 Neural Network Control Policy Verification With Persistent Adversarial Perturbation ICML 2020