24 citations · 32 across the 4 of their papers we have counts for
1 paper · 1 filter
Evan Hubinger
This paper analyzes and compares 11 different proposals for building safe advanced AI under the current machine learning paradigm, including major contenders such as iterated ampli…