upvote

points

by inigyou4 hours ago |
comments
upvoteby cassianoleal3 hours ago|
[-]
How do you qualify what makes a model "Mythos class", and how do you reliably test for it?
reply
upvoteby generalizations3 hours ago|
parent|
[-]
Presumably a deepswe benchmark, which IIRC puts GLM 5.2 between opus 4.8 and fable.
reply