Gemini 3 Deep Think update released, achieving SOTA on ARC-AGI-2, MMMU-Pro, and HLE
An updated version of Gemini 3 Deep Think has been released, achieving SOTA on ARC-AGI-2, MMMU-Pro, and HLE, and earning gold-medal-level results in physics and chemistry olympiads. In a tweet, Noam Shazeer said that the best way to solve hard problems is still to think about them carefully, and included a link for further reading.
Why read this
SelectedAI score 81
A Google employee announced the Gemini 3 Deep Think update, listing its SOTA performance on ARC-AGI-2, MMMU-Pro, and HLE, along with gold-medal-level results in olympiad competitions.
Topics
Body
An updated Gemini 3 Deep Think is out today: 📈 Achieves SOTA on ARC-AGI-2, MMMU-Pro, and HLE. 🥇Gold-medal level on Physics & Chemistry Olympiads. It turns out the best way to solve hard problems is still to think about them. Read more: https://t.co/PMmuyRq90B https://t.co/cxCAbcEz7q