We’ll initially prioritize people who have demonstrated an interest in probing language models for safety issues (see attached, for an example), whether via traditional research or great blogs and Twitter threads. If that’s you, please fill out the form!

→ View original post on X — @anthropicai,