On the Opportunities and Risks of Foundation Models
Center for Research on Foundation Models (CRFM)（基础模型研究中心）；Stanford Institute for Human-Centered Artificial Intelligence (HAI)（斯坦福大学以人为本人工智能研究院）；Stanford University（斯坦福大学）
Rishi Bommasani, Drew A. Hudson, Ehsan Adeli, Russ Altman, Simran Arora, Sydney von Arx, Michael S. Bernstein, Jeannette Bohg, Antoine Bosselut, Emma Brunskill,… 展开作者
Rishi Bommasani, Drew A. Hudson, Ehsan Adeli, Russ Altman, Simran Arora, Sydney von Arx, Michael S. Bernstein, Jeannette Bohg, Antoine Bosselut, Emma Brunskill, Erik Brynjolfsson, Shyamal Buch, Dallas Card, Rodrigo Castellon, Niladri Chatterji, Annie Chen, Kathleen Creel, Jared Quincy Davis, Dora Demszky, Chris Donahue, Moussa Doumbouya, Esin Durmus, Stefano Ermon, John Etchemendy, Kawin Ethayarajh, Li Fei-Fei, Chelsea Finn, Trevor Gale, Lauren Gillespie, Karan Goel, Noah Goodman, Shelby Grossman, Neel Guha, Tatsunori Hashimoto, Peter Henderson, John Hewitt, Daniel E. Ho, Jenny Hong, Kyle Hsu, Jing Huang, Thomas Icard, Saahil Jain, Dan Jurafsky, Pratyusha Kalluri, Siddharth Karamcheti, Geoff Keeling, Fereshte Khani, Omar Khattab, Pang Wei Koh, Mark Krass, Ranjay Krishna, Rohith Kuditipudi, Ananya Kumar, Faisal Ladhak, Mina Lee, Tony Lee, Jure Leskovec, Isabelle Levent, Xiang Lisa Li, Xuechen Li, Tengyu Ma, Ali Malik, Christopher D. Manning, Suvir Mirchandani, Eric Mitchell, Zanele Munyikwa, Suraj Nair, Avanika Narayan, Deepak Narayanan, Ben Newman, Allen Nie, Juan Carlos Niebles, Hamed Nilforoshan, Julian Nyarko, Giray Ogut, Laurel Orr, Isabel Papadimitriou, Joon Sung Park, Chris Piech, Eva Portelance, Christopher Potts, Aditi Raghunathan, Rob Reich, Hongyu Ren, Frieda Rong, Yusuf Roohani, Camilo Ruiz, Jack Ryan, Christopher Ré, Dorsa Sadigh, Shiori Sagawa, Keshav Santhanam, Andy Shih, Krishnan Srinivasan, Alex Tamkin, Rohan Taori, Armin W. Thomas, Florian Tramèr, Rose E. Wang, William Wang, Bohan Wu, Jiajun Wu, Yuhuai Wu, Sang Michael Xie, Michihiro Yasunaga, Jiaxuan You, Matei Zaharia, Michael Zhang, Tianyi Zhang, Xikun Zhang, Yuhui Zhang, Lucia Zheng, Kaitlyn Zhou, Percy Liang
备注：
Authored by the Center for Research on Foundation Models (CRFM) at the Stanford Institute for Human-Centered Artificial Intelligence (HAI). Report page with citation guidelines: https://crfm.stanford.edu/report.html
AI is undergoing a paradigm shift with the rise of models (e.g., BERT, DALL-E, GPT-3) that are trained on broad data at scale and are adaptable to a wide range of downstream tasks. We call these models foundation models to underscore their critically central yet incomplete character. This report provides a thorough account of the opportunities and risks of foundation models, ranging from their capabilities (e.g., language, vision, robotics, reasoning, human interaction) and technical principles(e.g., model architectures, training procedures, data, systems, security, evaluation, theory) to their applications (e.g., law, healthcare, education) and societal impact (e.g., inequity, misuse, economic and environmental impact, legal and ethical considerations). Though foundation models are based on standard deep learning and transfer learning, their scale results in new emergent capabilities,and their effectiveness across so many tasks incentivizes homogenization. Homogenization provides powerful leverage but demands caution, as the defects of the foundation model are inherited by all the adapted models downstream. Despite the impending widespread deployment of foundation models, we currently lack a clear understanding of how they work, when they fail, and what they are even capable of due to their emergent properties. To tackle these questions, we believe much of the critical research on foundation models will require deep interdisciplinary collaboration commensurate with their fundamentally sociotechnical nature.
Princeton University（普林斯顿大学）；Stanford University（斯坦福大学）；Massachusetts Institute of Technology（麻省理工学院）；Georgetown University（乔治城大学）；Center for Democracy and Technology（民主与技术中心）；Eleuther AI；Humane Intelligence；Brookings Institution（布鲁金斯学会）；Hugging Face；Australian National University（澳大利亚国立大学）；Open Source Initiative（开源倡议组织）；Institute for Advanced Study（高等研究院）；Meta；University of California, Berkeley（加州大学伯克利分校）；Mozilla AI
Risks and Opportunities of Open-Source Generative AI
University of Oxford（牛津大学）；MLCommons；Kenyon College（凯尼恩学院）；Center for Computation & Technology, Louisiana State University（路易斯安那州立大学计算与技术中心）；University of Texas at Austin（德克萨斯大学奥斯汀分校）；Institute for Technology & Society (ITS), Rio（里约技术与社会研究所）；St. John’s University（圣约翰大学）；University of California, Berkeley（加州大学伯克利分校）；University of Virginia（弗吉尼亚大学）；Luxembourg Institute of Science and Technology（卢森堡科学技术研究院）；University of Luxembourg（卢森堡大学）；University of Bath（巴斯大学）；National University Philippines（菲律宾国立大学）；ITESM（墨西哥蒙特雷科技大学）；Berklee College of Music（伯克利音乐学院）；University of Notre Dame（圣母大学）
AutoML in the Age of Large Language Models: Current Challenges, Future Opportunities and Risks
Institute of Artificial Intelligence, Leibniz University Hannover（汉诺威莱布尼茨大学人工智能学院）；Alma Mater Studiorum – University of Bologna（博洛尼亚大学）；German Research Center for Artificial Intelligence (DFKI)（德国人工智能研究中心）；L3S Research Center, Leibniz University Hannover（汉诺威莱布尼茨大学L3S研究中心）
Graph Foundation Models: Concepts, Opportunities and Challenges
School of Computer Science, Beijing University of Posts and Telecommunications（北京邮电大学计算机学院）；China Telecom Bestpay（中国电信翼支付）；School of Computing and Information Systems, Singapore Management University（新加坡管理大学计算与信息系统学院）；Lehigh University（利哈伊大学）；University of Illinois Chicago（伊利诺伊大学芝加哥分校）
Near to Mid-term Risks and Opportunities of Open-Source Generative AI
University of Oxford（牛津大学）；MLCommons；Kenyon College（凯尼恩学院）；Center for Computation & Technology, Louisiana State University（路易斯安那州立大学计算与技术中心）；Institute for Technology & Society (ITS), Rio（里约技术与社会研究所）；University of California, Berkeley（加州大学伯克利分校）；University of Virginia（弗吉尼亚大学）；Luxembourg Institute of Science and Technology（卢森堡科学与技术研究院）；University of Luxembourg（卢森堡大学）；University of Bath（巴斯大学）；National University Philippines（菲律宾国立大学）；ITESM（蒙特雷科技大学）；Berklee College of Music（伯克利音乐学院）；Bocconi University（博科尼大学）；University of Notre Dame（圣母大学）
An Investigation of Memorization Risk in Healthcare Foundation Models
MIT Broad Institute of MIT and Harvard（MIT Broad研究所（MIT和哈佛））；Vector Institute（向量研究所）；MIT；Chalmers University of Technology（查尔姆斯理工大学）；University of Gothenburg（哥德堡大学）；University of Toronto（多伦多大学）；University Health Network (UHN)（大学健康网络）；Worcester Polytechnic Institute（沃思维尔理工学院）；Computer Science Department（计算机科学系）
Decision-Oriented Uncertainty Quantification for Risk Control in Earth System Spatiotemporal Foundation Models
Vanderbilt University（范德堡大学）；City University of New York（纽约城市大学）；Yale University（耶鲁大学）；Wyze Inc.（Wyze公司）；Northeastern University（东北大学）；University of Southern California（南加州大学）；Southwest University（西南大学）
Protecting patient privacy in clinical foundation models: Technical and legal perspectives
Massachusetts Institute of Technology (MIT)（麻省理工学院（MIT））；The Broad Institute of MIT and Harvard（麻省理工学院与哈佛大学博德研究所）；Borealis AI；Stanford University（斯坦福大学）；Aalto University（阿尔托大学）；Max Planck Institute for Tax Law and Public Finance（马克斯·普朗克税法与公共财政研究所）；Stanford Law School（斯坦福法学院）；Harvard Law School（哈佛法学院）；Petrie-Flom Center for Health Law Policy, Biotechnology & Bioethics（皮里-弗洛姆健康法律政策、生物技术与生物伦理中心）
Open Foundation Models in Healthcare: Challenges, Paradoxes, and Opportunities with GenAI Driven Personalized Prescription
Qatar University（卡塔尔大学）；Birmingham City University（伯明翰城市大学）；Hamad Bin Khalifa University (HBKU)（哈马德·本·哈利法大学）；University of the West of England (UWE)（西英格兰大学）
Advancing Financial Risk Prediction Through Optimized LSTM Model Performance and Comparative Analysis
Columbia University（哥伦比亚大学）；Rutgers University（罗格斯大学）；University of Connecticut Business School（康涅狄格大学商学院）；Stevens Institute of Technological（史蒂文斯理工学院）
Tabular Foundation Models for Credit Risk Prediction: A Dedicated Review and Empirical Comparison
Faculty of Economics and Business, KU Leuven, Belgium（比利时库勒万大学经济与商业学院）；School of Business and Economics, Humboldt University of Berlin, Germany（德国洪堡大学商学院）；Department of Statistical and Actuarial Sciences, Western University, Canada（加拿大西部大学统计与精算科学系）；Department of Engineering Management, University of Antwerp, Belgium（比利时安特卫普大学工程管理系）；Business School, University of Edinburgh, United Kingdom（英国爱丁堡大学商学院）；Business School, University of Southampton, United Kingdom（英国南安普顿大学商学院）；School of Mathematical Sciences, University of Southampton, United Kingdom（英国南安普顿大学数学科学学院）；Department of Business Informatics and Operations Management, Ghent University, Belgium（比利时根特大学商业信息与运营管理系）；Department of Mathematics, University of Antwerp, Belgium（比利时安特卫普大学数学系）；Department of Mathematics, KU Leuven, Belgium（比利时库勒万大学数学系）
Foundation Models for Environmental Science: A Survey of Emerging Frontiers
University of Pittsburgh（匹兹堡大学）；University of Maryland（马里兰大学）；University of North Carolina at Chapel Hill（北卡罗来纳大学教堂山分校）；Lawrence Berkeley National Laboratory（劳伦斯伯克利国家实验室）
