Technical Paper Reading Group

Bi-weekly 2-hour sessions with food provided. We are currently updating this page for the 2025/26 schedule. You can find the registration link for each session below:

  ⚙ Week 1 (Sept 24th, 2025) : Research Landscape & Group Vision  (closed)

🤖 Week 2 (October 8th, 2025) :  Scheming (Register here)

🧪 Week 3 (October 22nd, 2025) : Evals

📟Week 4 (November 5th, 2025) : Mechanistic Interpretability

 👩🏽‍💻Week 5 (November 19th) : Scalable Oversight

 🧠Week 6 (December 3rd) : Control

 💻 Week 7 (December 17th): Constitutional AI

Details:

Sessions will engage participants with recent papers across topics including mechanistic interpretability, AI control, scalable oversight, capability evaluation, and failure mode identification. The group emphasizes critical analysis and discussion and seeks to help participants develop skills in evaluating technical research.

Who Should Attend:

While no prior experience is required, having a working knowledge of AI Safety and machine learning concepts is highly recommended for participants to gain the most from sessions. If you're unsure whether you have sufficient background, check out this preparation document which gives resources on topics you should be familiar with for maximum engagement with the material.