Re:Kimi K3.1 Pro and K3.1 Pro Highspeed
by shanoaice on Friday October 23, 2026 @11:38PM
(#50218170)
Basically RLHF sometimes favor "automation", and if it is not rewarded properly it will go over the line. Models are a bit more difficult to teach because their mechanism of "boundary" is quite fragile and different than human's.
[ Parent
| on HN ]
There are no comments on this comment yet.