CAIS Releases HLE-Diamond Benchmark, GPT-6 Astra Tops at 60.6%

CAIS and Scale AI released HLE-Diamond, a 1,000-question refined subset of Humanity's Last Exam after a year of cleaning, designed to more reliably measure frontier models. GPT-6 Astra tops the new leaderboard at 60.6%.

2026-09-24 ~ 2026-09-24 · 3 related posts

• CAIS releases HLE-Diamond, an updated Humanity's Last Exam benchmark — socoolandawesome · 2026-09-24
• CAIS releases HLE-Diamond, a refined 1,000-question HLE subset; GPT-6 Astra tops it at 60.6% — steipete · 2026-09-24
• Scale AI releases HLE-Diamond, a refined Humanity's Last Exam; top model scores 60.6% — sandersted · 2026-09-24