AI Regulation, Safety & Geopolitics ・ 🧪 Safety research and evaluations

Humanity's Last Exam: AI Benchmark Guide

CASRAI 🔗 Share link 11/07 15:24 UTC 🔗 Share link ITU, ISO and IEC explore international standards for AI safety and accountability — Tech Review Africa · 11/07 15:23 UTC

Humanity's Last Exam (HLE) is a 2,500-question frontier benchmark by CAIS and Scale AI. The guide compares model scores, evaluation steps and human expert baselines.

Read the original (CASRAI 🔗 Share link 11/07 15:24 UTC 🔗 Share link ITU, ISO and IEC explore international standards for AI safety and accountability — Tech Review Africa)

See it in the digest → ・ See this issue →