ai safety

Chaos with Keywords: Exposing Large Language Models Sycophancy to Misleading Keywords and Evaluating Defense Strategies ACL 2024 Characterising Interventions in Causal Games UAI 2024 Cognitive Overload: Jailbreaking Large Language Models with Overloaded Logical Thinking NAACL 2024 Compromesso! Italian Many-Shot Jailbreaks undermine the safety of Large Language Models ACL 2024 CRITIC: Large Language Models Can Self-Correct with Tool-Interactive Critiquing ICLR 2024 Data Efficient Paradigms for Personalized Assessment of Black-Box Taskable AI Systems AAAI 2024 Defending Against Alignment-Breaking Attacks via Robustly Aligned LLM ACL 2024 Discovering Agents (Abstract Reprint) AAAI 2024 Disinformation Capabilities of Large Language Models ACL 2024 Diversity-Aware Annotation for Conversational AI Safety COLING 2024 Do LLMs Know to Respect Copyright Notice? EMNLP 2024 Do-Not-Answer: Evaluating Safeguards in LLMs EACL 2024 Enhancing Large Language Models Against Inductive Instructions with Dual-critique Prompting NAACL 2024 Expectation Alignment: Handling Reward Misspecification in the Presence of Expectation Mismatch NeurIPS 2024 Flames: Benchmarking Value Alignment of LLMs in Chinese NAACL 2024 Fundamental Limitations of Alignment in Large Language Models ICML 2024 GPT-4 Is Too Smart To Be Safe: Stealthy Chat with LLMs via Cipher ICLR 2024 GPT-4 Jailbreaks Itself with Near-Perfect Success Using Self-Explanation EMNLP 2024 HonestLLM: Toward an Honest and Helpful Large Language Model NeurIPS 2024 How Johnny Can Persuade LLMs to Jailbreak Them: Rethinking Persuasion to Challenge AI Safety by Humanizing LLMs ACL 2024 How should Conversational Agent systems respond to sexual harassment? EACL 2024 Identifying the Risks of LM Agents with an LM-Emulated Sandbox ICLR 2024 Implications of Regulations on Large Generative AI Models in the Super-Election Year and the Impact on Disinformation COLING 2024 Improving Alignment and Robustness with Circuit Breakers NeurIPS 2024 INDICT: Code Generation with Internal Dialogues of Critiques for Both Security and Helpfulness NeurIPS 2024