Papers

∗ Equal contributors · † Core contributors

September 2026
Alignment forecasting: predicting misalignment from training data
Paper · Preprint
Chen Yueh-Han, Bruce W. Lee, Ilia Sucholutsky, Tomek Korbak
March 2026
Reasoning models struggle to control their chains of thought
Paper · ICML 2026
Chen Yueh-Han, Robert McCarthy, Bruce W. Lee, He He, Ian Kivlichan, Bowen Baker, Micah Carroll, Tomek Korbak
February 2026
Training agents to self-report misbehavior
Paper · COLM 2026
Bruce W. Lee, Chen Yueh-Han, Tomek Korbak
June 2025
Distillation robustifies unlearning
Paper · NeurIPS 2025 (Spotlight)
Bruce W. Lee^†, Addie Foote^†, Alex Infanger^†, Leni Shor^†, Harish Kamath^†, Jacob Goldman-Wetzler, Bryce Woodworth, Alex Cloud, Alexander Matt Turner
February 2025
Utility engineering: analyzing and controlling emergent value systems in AIs
Paper · NeurIPS 2025 (Spotlight)
Mantas Mazeika, Xuwang Yin, Rishub Tamirisa, Jaehyuk Lim, Bruce W. Lee, Richard Ren, Long Phan, Norman Mu, Adam Khoja, Oliver Zhang, Dan Hendrycks
September 2024
Programming refusal with conditional activation steering
Paper · ICLR 2025 (Spotlight)
Bruce W. Lee, Inkit Padhi, Karthikeyan Natesan Ramamurthy, Erik Miehling, Pierre Dognin, Manish Nagireddy, Amit Dhurandhar
August 2024
Inertia in moral and value judgments of large language models
Paper · ACL 2026
Bruce W. Lee, Yeongheon Lee, Hyunsoo Cho
February 2024
Language models don’t learn the physical manifestation of language
Paper · ACL 2024
Bruce W. Lee, JaeHyuk Lim
October 2023
Instruction tuning with human curriculum
Paper · NAACL 2024
Bruce W. Lee^∗, Hyunsoo Cho^∗, Kang Min Yoo
February 2023
Prompt-based learning for text readability assessment
Paper · EACL 2023 (Findings)
Bruce W. Lee, Jason Hyung-Jong Lee
September 2021
Pushing on text readability assessment: a transformer meets handcrafted linguistic features
Paper · EMNLP 2021
Bruce W. Lee, Yoo Sung Jang, Jason Hyung-Jong Lee