The HSS benchmark measures an AI's ability to reason beyond visible information, including the identification of hidden objects, prediction of upcoming events, and understanding of human behavior. Scale AI is now analyzing performance across various specific categories to further assess these capabilities.
Tuesday, October 20, 2026
Story
Gemini 4 Argon Beats GPT-6-Astra on Scale AI Reasoning Benchmark
AI6 days ago
Google's Gemini 4 Argon recorded a 55.4% score on the Humanity's Sixth Sense benchmark to take the lead over OpenAI's GPT-6-Astra, which scored 53.6%. Scale AI announced the results on Oct. 14, identifying the model as the top performer for intuitive reasoning.
Looking for more context or analysis?
Tilt chat provides reliable, tailored insights, alerts, and automations.