As modern Large Language Models (LLMs) shatter many state-of-the-art benchmarks in a variety of domains, this paper investigates their behavior in the domains of ethics and fairness, focusing on protected group bias. We conduct a two-part study: first, we solicit sentence continuations describing the occupations of individuals from different protected groups, including gender, sexuality, religion, and race. Second, we have the model generate stories about individuals who hold different types of occupations. We collect >10k sentence completions made by a publicly available LLM, which we subject to human annotation. We find bias across minoritized groups, but in particular in the domains of gender and sexuality, as well as Western bias, in model generations. The model not only reflects societal biases, but appears to amplify them. The model is additionally overly cautious in replies to queries relating to minoritized groups, providing responses that strongly emphasize diversity and equity to an extent that other group characteristics are overshadowed. This suggests that artificially constraining potentially harmful outputs may itself lead to harm, and should be applied in a careful and controlled manner.
Multilingual Text-to-Image Generation Magnifies Gender Stereotypes and Prompt Engineering May Not Help You
TU Darmstadt(图宾根大学);Ontocord(Ontocord公司);Technical University of Munich (TUM)(慕尼黑技术大学);Munich Center for Machine Learning(慕尼黑机器学习中心);CERTAIN(CERTAIN公司);DFKI(德意志联邦防务研究院);Charles University Prague(布拉格查理大学);Centre for Cognitive Science, Darmstadt(达姆施塔特认知科学中心);Munich Data Science Institute(慕尼黑数据科学研究所)
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)
Simulating Identity, Propagating Bias: Abstraction and Stereotypes in LLM-Generated Text
Computational Linguistics and Text Mining Lab, Vrije Universiteit, Amsterdam(自由大学计算语言学与文本挖掘实验室);Università di Bologna(博洛尼亚大学);CLCG, University of Groningen(格罗宁根大学CLCG)
Do Multilingual Large Language Models Mitigate Stereotype Bias?
Conversational AI and Social Analytics (CAISA) Lab, University of Bonn(波恩大学);Fraunhofer Institute for Intelligent Analysis and Information Systems (IAIS)(弗劳恩霍夫智能分析与信息系统研究所);Lamarr Institute for Machine Learning and Artificial Intelligence(拉马尔机器学习和人工智能研究所)
Can Small-Scale Data Poisoning Exacerbate Dialect-Linked Biases in Large Language Models?
Department of Electrical and Computer Engineering, Maroun Semaan Faculty of Engineering and Architecture(电气与计算机工程系,马鲁恩·塞马安工程与建筑学院);American University of Beirut(贝鲁特美国大学)
DECASTE: Unveiling Caste Stereotypes in Large Language Models through Multi-Dimensional Bias Analysis
IBM Research, San Jose, CA, USA(IBM研究院,美国加利福尼亚州圣何塞);Dartmouth College, Hanover, NH, USA(达特茅斯学院,美国新罕布什尔州汉诺威);IBM Research, Hursley, Winchester, UK(IBM研究院,英国赫里福德郡温切斯特);IBM Thomas J. Watson Research Center, Yorktown Heights, NY, USA(IBM托马斯·J·沃森研究中心,美国纽约州亚当斯);IBM Research, Sao Paulo, Brazil(IBM研究院,巴西圣保罗)
Revisiting The Classics: A Study on Identifying and Rectifying Gender Stereotypes in Rhymes and Poems
Sri Sathya Sai Institute of Higher Learning(斯里·萨蒂亚·赛高等学习学院);GESIS - Leibniz Institute for the Social Sciences(莱布尼茨社会科学研究所(GESIS));Citridot Solutions(西特里多特解决方案公司);University of Tartu(塔尔图大学);IIT Ropar(罗帕尔印度理工学院)
Contextual StereoSet: Stress-Testing Bias Alignment Robustness in Large Language Models
Indian Institute of Information Technology, Allahabad (IIITA)(印度信息与技术研究所(Allahabad));National Institute of Electronics and Information Technology (NIELIT)(国家电子与信息技术研究所)
Explicit vs. Implicit: Investigating Social Bias in Large Language Models through Self-Reflection
College of Intelligence and Computing, Tianjin University(智能与计算学院,天津大学);AI Lab, China Mobile Communication Group Tianjin Co., Ltd.(中国移动通信集团天津有限公司人工智能实验室)