AI Safety Evals - Paper Reading Club

​Mark Keavney will present Training large language models on narrow tasks can lead to broad misalignment, one of the foundational studies of emergent misalignment, a counterintuitive and fascinating finding in alignment research.

​​Every week, someone will present for up to 20 minutes followed by 40 minutes of discussion. RSVP to join, sign up to present, or contact us at [email protected] with questions. Everyone is welcome!