Fabien Roger

According to our database, Fabien Roger authored at least 8 papers between 2023 and 2024.

Collaborative distances:

• Dijkstra number of five.
• Erdős number of four.

2024

Language Models Are Better Than Humans at Next-token Prediction.

Buck Shlegeris, Fabien Roger, Lawrence Chan, Euan McLean

Trans. Mach. Learn. Res., 2024

Alignment faking in large language models.

Ryan Greenblatt, Carson Denison, Benjamin Wright, Fabien Roger, Monte MacDiarmid, Samuel Marks, Johannes Treutlein, Tim Belonax, Jack Chen, David Duvenaud, Akbir Khan, Julian Michael, Sören Mindermann, Ethan Perez, Linda Petrini, Jonathan Uesato, Jared Kaplan, Buck Shlegeris, Samuel R. Bowman, Evan Hubinger

CoRR, 2024

Do Unlearning Methods Remove Information from Language Model Weights?

Aghyad Deeb, Fabien Roger

CoRR, 2024

Stress-Testing Capability Elicitation With Password-Locked Models.

Ryan Greenblatt, Fabien Roger, Dmitrii Krasheninnikov, David Krueger

Proceedings of the Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information Processing Systems 2024, 2024

AI Control: Improving Safety Despite Intentional Subversion.

Ryan Greenblatt, Buck Shlegeris, Kshitij Sachan, Fabien Roger

Proceedings of the Forty-first International Conference on Machine Learning, 2024

2023

Preventing Language Models From Hiding Their Reasoning.

Fabien Roger, Ryan Greenblatt

CoRR, 2023

Measurement Tampering Detection Benchmark.

Fabien Roger, Ryan Greenblatt, Max Nadeau, Buck Shlegeris, Nate Thomas

CoRR, 2023

Large Language Models Sometimes Generate Purely Negatively-Reinforced Text.

Fabien Roger

CoRR, 2023