Max Lamparth Headshot 2024

Max Lamparth

  • Post-doctoral Fellow

Biography

Max is a returning postdoctoral fellow at the Center for International Security and Cooperation, the Stanford Center for AI Safety, and the Stanford Existential Risks Initiative. With his technical research, Max wants to make AI systems more secure and safe to use. Specifically, he his focussing on how to improve the ethical behavior of language models, how to make their inner workings more interpretable, and how to increase their robustness against misuses. 

Max also created and teaches a new course at Stanford: “CS120: Introduction to AI Safety”. His research and teaching is advised by Prof. Clark Barrett, Prof. Steve Luby, and Prof. Paul Edwards. Previously, Max received his Ph.D. in August 2023 from the Technical University of Munich and previously a B.Sc. and M.Sc. from the Ruprecht Karl University of Heidelberg.

publications

Journal Articles
June 2024

Analyzing And Editing Inner Mechanisms of Backdoored Language Models

Author(s)
cover link Analyzing And Editing Inner Mechanisms of Backdoored Language Models
Journal Articles
June 2024

Human vs. Machine: Behavioral Differences Between Expert Humans and Language Models in Wargame Simulations

Author(s)
cover link Human vs. Machine: Behavioral Differences Between Expert Humans and Language Models in Wargame Simulations
Journal Articles
May 2024

Markovian Agents for Informative Language Modeling

Author(s)
cover link Markovian Agents for Informative Language Modeling