Large language models (LLMs) have exploded in popularity in the past few years and have achieved undeniably impressive results on benchmarks as varied as question answering and text summarization. We provide a simple new prompting strategy that leads to yet another supposedly "super-human" result, this time outperforming humans at common sense ethical reasoning (as measured by accuracy on a subset of the ETHICS dataset). Unfortunately, we find that relying on average performance to judge capabilities can be highly misleading. LLM errors differ systematically from human errors in ways that make it easy to craft adversarial examples, or even perturb existing examples to flip the output label. We also observe signs of inverse scaling with model size on some examples, and show that prompting models to "explain their reasoning" often leads to alarming justifications of unethical actions. Our results highlight how human-like performance does not necessarily imply human-like understanding or reasoning.
SafeMed-R1: Clinician-Audited Safety and Ethics Alignment for Medical Large Language Models
Shanghai Artificial Intelligence Laboratory(上海人工智能实验室);Joint Laboratory of Biomedical Artificial Intelligence(生物医学人工智能联合实验室);Shanghai Institute of Infectious Disease and Biosecurity(上海传染病与生物安全研究院);Shanghai Health Development Research Center (Shanghai Medical Information Center)(上海健康发展战略研究中心(上海医疗信息中心));University of Washington(华盛顿大学);Department of Eye and Vision Sciences, University of Liverpool(利物浦大学眼科与视觉科学系);Liverpool Centre for Cardiovascular Science, University of Liverpool(利物浦大学心血管科学中心);School of Computer Science and Technology, Tongji University(同济大学计算机科学与技术学院)
Between a Rock and a Hard Place: The Tension Between Ethical Reasoning and Safety Alignment in LLMs
Department of Computer Science and Technology, Institute for AI, BNRist, Tsinghua University(计算机科学与技术系,人工智能研究院,BNRist,清华大学);IDG/McGovern Institute for Brain Research, Tsinghua University(IDG/麦克戈文脑科学研究院,清华大学);Chinese Institute for Brain Research (CIBR)(中国脑科学研究院(CIBR));Shanghai Jiao Tong University(上海交通大学);ByteDance(字节跳动)
The Moral Consistency Pipeline: Continuous Ethical Evaluation for Large Language Models
SWAT Laboratory, Polytechnique Montréal(蒙特利尔大学SWAT实验室);Concordia Institute for Information Systems Engineering, Concordia University(Concordia大学信息系统工程研究所)
Revealing the Intrinsic Ethical Vulnerability of Aligned Large Language Models
Department of Electrical and Electronic Engineering(电子与电气工程系);The Hong Kong Polytechnic University(香港理工大学);School of Electronics and Information(电子与信息学院);Northwestern Polytechnical University(西北工业大学)
PromptGuard: An Orchestrated Prompting Framework for Principled Synthetic Text Generation for Vulnerable Populations using LLMs with Enhanced Safety, Fairness, and Controllability
Posts and Telecommunications Institute of Technology(邮电技术研究所);Hanoi Architectural University(河内建筑大学)
Analyzing the Safety of Japanese Large Language Models in Stereotype-Triggering Prompts
Graduate School of Science and Technology, University of Tsukuba(筑波大学科学技术研究科);Institute of Systems and Information Engineering, University of Tsukuba(筑波大学系统与信息工程学院);Department of Electronics, Information and Bioengineering, Politecnico di Milano(米兰理工大学电子、信息与生物工程系)
FinSafetyBench: Evaluating LLM Safety in Real-World Financial Scenarios
Shanghai University of Finance and Economics(上海财经大学);Beihang University(北航);Southern University of Science and Technology(南方科技大学);MoE Key Laboratory of Interdisciplinary Research of Computation and Economics(计算与经济学交叉学科研究重点实验室)