cooperation and disincentivizing high-risk approach
Foundational Challenges in Assuring Alignment and Safety of Large Language Models
Anwar, Usman, Saparov, Abulhair, Rando, Javier, Paleka, Daniel, Turpin, Miles, Hase, Peter, Lubana, Ekdeep Singh, Jenner, Erik, Casper, Stephen, Sourbut, Oliver, Edelman, Benjamin L., Zhang, Zhaowei, Günther, Mario, Korinek, Anton, Hernandez-Orallo, Jose, Hammond, Lewis, Bigelow, Eric, Pan, Alexander, Langosco, Lauro, Korbak, Tomasz, Zhang, Heidi, Zhong, Ruiqi, hÉigeartaigh, Seán Ó, Recchia, Gabriel, Corsi, Giulio, Chan, Alan, Anderljung, Markus, Edwards, Lilian, Bengio, Yoshua, Chen, Danqi, Albanie, Samuel, Maharaj, Tegan, Foerster, Jakob, Tramer, Florian, He, He, Kasirzadeh, Atoosa, Choi, Yejin, Krueger, David
This work identifies 18 foundational challenges in assuring the alignment and safety of large language models (LLMs). These challenges are organized into three different categories: scientific understanding of LLMs, development and deployment methods, and sociotechnical challenges. Based on the identified challenges, we pose $200+$ concrete research questions.