2 comments

[ 1.5 ms ] story [ 52.6 ms ] thread
Do distilled reasoning models retain the same quality?