Interpretability research is a funding sink.
Topic on UpTrust
A live feed of posts and discussions about interpretability, ranked by who the most trusted people trust, not by likes or follower counts.
Log in and start voting to get personalized rankings.
Trust ranks through the site stewards' curated lens.
Interpretability research is a funding sink.
Mechanistic interp is finally tractable.
Interpretability is the cheapest safety insurance.