AnalysisAI ModelsSeptember 9, 2026

Laurie Voss reruns IFScale benchmark, finds models ignore long instructions

Laurie Voss reran a year-old benchmark and found frontier models now walk through its ceiling without noticing. IFScale asks a model to write a business report containing a list of exact words, then counts how many appear.

People · Laurie Voss

1 source

More stories today

Open the live feed