2 comments

[ 5.4 ms ] story [ 21.9 ms ] thread
Do distilled reasoning models retain the same quality?