The literature and multiple experts point to many potential risks from large language models (LLMs), but there are still very few direct measurements of the actual harms posed. AI risk assessment has so far focused on measuring the models' capabilities, but the capabilities of models are only indicators of risk, not measures of risk. Better modeling and quantification of AI risk scenarios can help bridge this disconnect and link the capabilities of LLMs to tangible real-world harm. This paper makes an early contribution to this field by demonstrating how existing AI benchmarks can be used to facilitate the creation of risk estimates. We describe the results of a pilot study in which experts use information from Cybench, an AI benchmark, to generate probability estimates. We show that the methodology seems promising for this purpose, while noting improvements that can be made to further strengthen its application in quantitative AI risk assessment.
IDEA2: Expert-in-the-loop competency question elicitation for collaborative ontology engineering
School of Computer Science and Informatics, University of Liverpool(利物浦大学计算机科学与信息学院);College of Computer Science and Engineering, Taibah University(塔伊巴大学计算机科学与工程学院);Unilever Plc., Materials Innovation Factory, University of Liverpool(联合利华公司,利物浦大学材料创新工厂)
Beyond Value Elicitation: Towards Moral Profiles in Early Requirements Engineering via Role-Playing Games and Anthropologist LLMs
Computer Science Area, Gran Sasso Science Institute(格兰萨索科学研究所计算机科学部门);Department of Computer Science, University of Pisa(比萨大学计算机科学系);Department of Humanities, University of L’Aquila(拉奎拉大学人文学系)
Preference-Guided Planning: An Active Elicitation Approach
University of Texas, Dallas(得克萨斯大学达拉斯分校);Indiana University, Bloomington(印第安纳大学伯明顿分校);Washington State University(华盛顿州立大学);University of Pennsylvania(宾夕法尼亚大学)
Auditing a KB Elicitation of Frontier LLM Knowledge: A Multi-dimensional Analysis of GPTKB v1.5
University of Tübingen, Germany(图宾根大学);ScaDS.AI Dresden/Leipzig & TU Dresden, Germany(ScaDS.AI 德累斯顿/莱比希 & 德累斯顿技术大学);VNU University of Engineering and Technology, Hanoi, Vietnam(越南河内工程与技术大学)
Large language models as synthetic clinical experts to inform longitudinal rare-disease modeling
University of Freiburg(弗莱堡大学);Freiburg Center for Data Analysis, Modeling and AI(弗莱堡数据分析、建模与人工智能中心);University of Bonn(波恩大学);Bonn Center for Mathematical Life Sciences(波恩数学生命科学中心);Life and Medical Sciences (LIMES) Institute(生命与医学科学研究所);Centre for Integrative Biological Signalling Studies(整合生物信号研究中心)
Improving drug sensitivity predictions in precision medicine through active expert knowledge elicitation
Helsinki Institute for Information Technology HIIT(赫尔辛基信息技术研究所(HIIT));Aalto University(阿尔托大学);Institute for Molecular Medicine Finland (FIMM)(芬兰分子医学研究所(FIMM));University of Helsinki(赫尔辛基大学)
Collective intelligence in science: direct elicitation of diverse information from experts with unknown information structure
Swiss Re Institute(瑞士再保险学院);St. Petersburg State University(圣彼得堡国立大学);St. Petersburg Department of V. A. Steklov Institute of Mathematics of the Russian Academy of Sciences(俄罗斯科学院圣彼得堡维克托·阿列克谢耶维奇·斯捷克洛夫数学研究所圣彼得堡分所);HSE University(俄罗斯高等经济大学)
Estimating Item Difficulty with Large Language Models as Experts
Department of Methodology and Statistics, Tilburg University(蒂尔堡大学方法学与统计学系);Smart Business Technologies(智能商务技术公司);Department of Psychological Methods, University of Amsterdam(阿姆斯特丹大学心理方法系);Prowise Learn, Amsterdam(Prowise Learn公司,阿姆斯特丹);Department of Information and Computing Sciences, Utrecht University(乌得勒支大学信息与计算科学系)
The Tournament Tree Method for preference elicitation in Multi-criteria decision-making
Department of Mathematics, Universidad de Ja \'e n, 23071 Ja \'e n, Spain;Department of Computer Science, Universidad de Ja \'e n, 23071 Ja \'e n, Spain
Logic-Q: Improving Deep Reinforcement Learning-based Quantitative Trading via Program Sketch-based Tuning
Nanyang Technological University(南洋理工大学);Hong Kong Polytechnic University(香港理工大学);Chinese University of Hong Kong(香港中文大学);Singapore Management University(新加坡管理大学);Peng Cheng Lab(鹏城实验室)