Global AI News Aggregator
About
By
–
Gemini 3 Pro benchmarks are wild – Humanity’s Last Exam: 37.5% – ARC-AGI-2: 31.1% True SOTA
→ View original post on X — @testingcatalog