2 comments

[ 3.3 ms ] story [ 16.3 ms ] thread
Do distilled reasoning models retain the same quality?