I evaluated GLM 5.2 against the frontier on tasks from real repos (opens in new tab)

(stet.sh)

2 pointsbisonbear7h ago1 comments

1 comments

1 comments · 1 top-level

Why did you use GLM 5.2 on medium when you used all the other models on high or xhigh?

j / k navigate · click thread line to collapse