Presented by
BlueDot Impact
AI Safety Evals - Paper Reading Club - Juliana Grant presents on CoT Monitorability
Registration
Past Event
About Event
Juliana Grant (member) presents the A Pragmatic Way to Measure Chain-of-Thought Monitorability by Emmons, Zimmermann, et al. Paper 🔗.
This paper proposes a method to measure the monitorability of Chain-of-Thought reasoning in AI models through metrics of legibility and coverage, providing a tool for developers to assess the impact of design decisions on monitorability.
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 evalsreadinggroup@gmail.com with questions. Everyone is welcome!
Presented by
BlueDot Impact