AI Safety Researcher
Rowan is a co-author on research into overthinking in reasoning models, investigating how large language models allocate computational effort.
Collaborations: