← Core AI
Artificial Intelligence › Core AI ›

AI Safety

2972 directly classified papers

Papers per year

Papers

Neural Abstractions NIPS 2022
One-shot Neural Backdoor Erasing via Adversarial Weight Masking NIPS 2022
Sleeper Agent: Scalable Hidden Trigger Backdoors for Neural Networks Trained from Scratch NIPS 2022
Pre-activation Distributions Expose Backdoor Neurons NIPS 2022
When Adversarial Training Meets Vision Transformers: Recipes from Training to Architecture NIPS 2022
Lexicographic Multi-Objective Reinforcement Learning IJCAI 2022
Neural Theory-of-Mind? On the Limits of Social Intelligence in Large LMs EMNLP 2022
SafeText: A Benchmark for Exploring Physical Safety in Language Models EMNLP 2022
Verification and Monitoring for First-Order LTL with Persistence-Preserving Quantification over Finite and Infinite Traces IJCAI 2022
A Universal Identity Backdoor Attack against Speaker Verification based on Siamese Network INTERSPEECH 2022
Towards Practical Certifiable Patch Defense With Vision Transformer CVPR 2022
BppAttack: Stealthy and Efficient Trojan Attacks Against Deep Neural Networks via Image Quantization and Contrastive Adversarial Learning CVPR 2022
Certified Patch Robustness via Smoothed Vision Transformers CVPR 2022
Quarantine: Sparsity Can Uncover the Trojan Attack Trigger for Free CVPR 2022
Fingerprinting Deep Neural Networks Globally via Universal Adversarial Perturbations CVPR 2022
Practical Evaluation of Adversarial Robustness via Adaptive Auto Attack CVPR 2022
Understanding the Limits of Poisoning Attacks in Episodic Reinforcement Learning IJCAI 2022
On the Adversarial Robustness of Causal Algorithmic Recourse ICML 2022
Not All Poisons are Created Equal: Robust Training against Data Poisoning ICML 2022
Fast and Reliable Evaluation of Adversarial Robustness with Minimum-Margin Attack ICML 2022
Certified Robustness Against Natural Language Attacks by Causal Intervention ICML 2022
An Equivalence Between Data Poisoning and Byzantine Gradient Attacks ICML 2022
Path-Specific Objectives for Safer Agent Incentives AAAI 2022
ROSE: Robust Selective Fine-tuning for Pre-trained Language Models EMNLP 2022
GoTube: Scalable Statistical Verification of Continuous-Depth Models AAAI 2022