Part of this work was done during a residency at @Iliad_research. They run really great fellowships and courses in Berkeley and London. Highly recommend!
GPT-5.5-Cyber is our most capable cyber model yet, designed for advanced, authorized defensive work: tracing vulnerable code, validating issues, developing patches, and preparing evidence for human review.
We're hiring at Iliad!
Iliad is a small and quickly growing team building the field of foundational AI alignment. We organize conferences, incubate an Alignment journal and new alignment organizations, run a fellowship, and teach the foundations in the Iliad Intensive. 🧵
We just released the full course materials of the Iliad Intensive — a month-long, full-time AI alignment course for mathematicians, physicists, and theoretical computer scientists.
~20 contributors, 19 modules, at a depth that doesn't exist elsewhere for most of these topics. 🧵
Iliad is still searching for a director of studies to coordinate the future intensive programs and scale them up to monthly iterations! Please apply below.
1) Love teaching and operations? Strong background in math, physics, or CS? Care about AI alignment?
We're hiring a Director of Studies for the Iliad Intensive — a quickly growing leading research course on foundational AI alignment. 🧵
We're also incubating an academic journal for AI alignment journal: Rapid peer-review of foundational research that the current publication ecosystem underserves. Learn more: https://t.co/55ynb8nhUT