SemiAnalysis Weekly

Jordan Nanos, Doug O'Laughlin

1 episode

• 

  Ep. 042 - GDPx: The Benchmark for Frontier AI

  • SemiAnalysis launched GDPx, a frontier AI benchmark spanning 11 domains and built with 14 specialized data partners.
  • On GDPx, frontier models scored 0% on finance under binary grading except Carbon, while partial grading put finance in the 50s and legal in the 30s or 40s.… more

  3 Nov · 52 min