Report
Gemini 4 Argon tops Humanity's Sixth Sense benchmark with 55.4%
Scale AI Labs says GPT-6-Astra scored 53.6%, trailing Gemini 4 Argon on its HSS benchmark.
TLDR
Scale AI Labs says Gemini 4 Argon took the top spot on its Humanity's Sixth Sense (HSS) benchmark with 55.4%, ahead of GPT-6-Astra at 53.6%. The benchmark tests AI reasoning beyond what's visible, including hidden objects, what might happen next and human behavior.
Combined views
—
2 Sources, first seen ago
