Buck Shlegeris

Principal Researcher, Redwood Research

Joined

September 2022

Names

Buck Shlegeris (Preferred)

Emails

****@rdwrs.com (Confirmed)

Personal Links

Homepage

Google Scholar

DBLP

Career & Education History

Principal Researcher

Redwood Research (rdwrs.com)

2021 – Present

Undergrad student

Australian National University (anu.edu.au)

2012 – 2014

Advisors, Relations & Conflicts

Coworker

Ryan Greenblatt

2021 – Present

Expertise

AI control

Present

deep learning

2019 – Present

AI safety

2017 – Present

Publications

Games for AI Control: Models of Safety Evaluations of AI Deployment Protocols
Charlie Griffin, Louis Thomson, Buck Shlegeris, Alessandro Abate
- ICLR 2026 Conference Withdrawn Submission
- Readers: Everyone
Subversion Strategy Eval: Can LLMs statelessly strategize to subvert control protocols?
Alex Troy Mallen, Charlie Griffin, Misha Wagner, Alessandro Abate, Buck Shlegeris
- Submitted to ICLR 2026
- Readers: Everyone
Subversion Strategy Eval: Evaluating AI’s stateless strategic capabilities against control protocols
Alex Troy Mallen, Charlie Griffin, Alessandro Abate, Buck Shlegeris
- airrworkshop OralandPoster
- Readers: Everyone
Shell Games: Control Protocols for Adversarial AI Agents
Aryan Bhatt, Cody Rushing, Adam Kaufman, Vasil Georgiev, Tyler Tracy, Akbir Khan, Buck Shlegeris
- ICLR 2025 Conference Withdrawn Submission
- Readers: Everyone
Adaptive Deployment of Untrusted LLMs Reduces Distributed Threats
Jiaxin Wen, Vivek Hebbar, Caleb Larson, Aryan Bhatt, Ansh Radhakrishnan, Mrinank Sharma, Henry Sleight, Shi Feng, He He, Ethan Perez, Buck Shlegeris, Akbir Khan
- ICLR 2025 Poster
- Readers: Everyone
Games for AI-Control: Models of Safety Evaluations of AI Deployment Protocols
Charlie Griffin, Buck Shlegeris, Alessandro Abate
- ICML 2024 TiFA Workshop
- Readers: Everyone
Language Models Are Better Than Humans at Next-token Prediction
Buck Shlegeris, Fabien Roger, Lawrence Chan, Euan McLean
- Accepted by TMLR
- Readers: Everyone
AI Control: Improving Safety Despite Intentional Subversion
Ryan Greenblatt, Buck Shlegeris, Kshitij Sachan, Fabien Roger
- ICML 2024 Oral
- Readers: Everyone
Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training
Evan Hubinger, Carson Denison, Jesse Mu, Mike Lambert, Meg Tong, Monte MacDiarmid, Tamera Lanham, Daniel M. Ziegler, Tim Maxwell, Newton Cheng, Adam S. Jermyn, Amanda Askell, Ansh Radhakrishnan, Cem Anil, David Duvenaud, Deep Ganguli, Fazl Barez, Jack Clark, Kamal Ndousse, Kshitij Sachan et al. (19 additional authors not shown)
- CoRR 2024
- Readers: Everyone
Language Models Are Better Than Humans at Next-token Prediction
Buck Shlegeris, Fabien Roger, Lawrence Chan, Euan McLean
- Trans. Mach. Learn. Res. 2024
- Readers: Everyone

View all 24 publications

Co-Authors

View all 80 co-authors