New models get instantly torn apart by 101 benchmarks: 'Cruella d'eval'

erhannah · x · 2026-10-16

A witty AI-culture joke: whenever a trendy new model drops, it is immediately butchered by 101 benchmarks — punning on Cruella de Vil of 101 Dalmatians to mock today's benchmark-saturated eval culture.

Original post →

More from Fun

Fun channel →