AI Safety

CRITIC: Large Language Models Can Self-Correct with Tool-Interactive Critiquing ICLR 2024 Cross-Task Defense: Instruction-Tuning LLMs for Content Safety NAACL 2024 CROWD: Certified Robustness via Weight Distribution for Smoothed Classifiers against Backdoor Attack EMNLP 2024 CR-UTP: Certified Robustness against Universal Text Perturbations on Large Language Models ACL 2024 CTIBench: A Benchmark for Evaluating LLMs in Cyber Threat Intelligence NeurIPS 2024 Curiosity-driven Red-teaming for Large Language Models ICLR 2024 DA3: A Distribution-Aware Adversarial Attack against Language Models EMNLP 2024 DALD: Improving Logits-based Detector without Logits from Black-box LLMs NeurIPS 2024 DarkFed: A Data-Free Backdoor Attack in Federated Learning IJCAI 2024 Data Advisor: Dynamic Data Curation for Safety Alignment of Large Language Models EMNLP 2024 Data-Driven Discovery of Design Specifications (Student Abstract) AAAI 2024 Data Efficient Paradigms for Personalized Assessment of Black-Box Taskable AI Systems AAAI 2024 Data Free Backdoor Attacks NeurIPS 2024 Data-Free Hard-Label Robustness Stealing Attack AAAI 2024 Dataset and Lessons Learned from the 2024 SaTML LLM Capture-the-Flag Competition NeurIPS 2024 Decision-Making Behavior Evaluation Framework for LLMs under Uncertain Context NeurIPS 2024 DeepBern-Nets: Taming the Complexity of Certifying Neural Networks Using Bernstein Polynomial Activations and Precise Bound Propagation AAAI 2024 DeepSaDe: Learning Neural Networks That Guarantee Domain Constraint Satisfaction AAAI 2024 Defeasible Normative Reasoning: A Proof-Theoretic Integration of Logical Argumentation AAAI 2024 Defending Against Alignment-Breaking Attacks via Robustly Aligned LLM ACL 2024 Defending Against Backdoor Attacks by Layer-wise Feature Analysis (Extended Abstract) IJCAI 2024 Defending Against Weight-Poisoning Backdoor Attacks for Parameter-Efficient Fine-Tuning NAACL 2024 Defending Jailbreak Prompts via In-Context Adversarial Game EMNLP 2024 Defending Large Language Models Against Jailbreak Attacks via Layer-specific Editing EMNLP 2024 Defending Large Language Models Against Jailbreaking Attacks Through Goal Prioritization ACL 2024