170 citations · 344 across the 3 of their papers we have counts for
3 papers
Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned
Deep Ganguli, Liane Lovitt, Jackson Kernion +33
We describe our early efforts to red team language models in order to simultaneously discover, measure, and attempt to reduce their potentially harmful outputs. We make three main…
Language Models (Mostly) Know What They Know
Saurav Kadavath, Tom Conerly, Amanda Askell +33
We study whether language models can evaluate the validity of their own claims and predict which questions they will be able to answer correctly. We first show that larger models a…
Moulting Black Holes
Iosif Bena, Borun D. Chowdhury, Jan de Boer +2
We find a family of novel supersymmetric phases of the D1-D5 CFT, which in certain ranges of charges have more entropy than all known ensembles. We also find bulk BPS configuration…