This year was a major paradigm shift, where we can solve problems end to end in natural language. With novel reinforcement learning techniques, we are able to train an advanced Gemini model on multi-step reasoning proof data, which advances the model's capabilities in terms of
MACHINE LEARNING
-
AI Achieves IMO Gold Medal via Token Space Search
By
–
Officially validated IMO gold medal, purely via search in token space, achieved in 4.5 hrs (unclear at what compute cost). The solutions read nicely as well
-
Gemini Deep Think Achieves Advanced Reasoning Capabilities
By
–
We achieved this year’s impressive result using an advanced version of Gemini Deep Think (an enhanced reasoning mode for complex problems). Our model operated end-to-end in natural language, producing rigorous mathematical proofs directly from the official problem descriptions –
-
Gemini Achieves Gold Medal Standard at International Mathematical Olympiad
By
–
Official results are in – Gemini achieved gold-medal level in the International Mathematical Olympiad! 🏆 An advanced version was able to solve 5 out of 6 problems. Incredible progress – huge congrats to @lmthang and the team! deepmind.google/discover/blo…
→ View original post on X — @goodfellow_ian, 2025-07-21 16:47 UTC
-
FourCastNet: First High-Resolution AI Weather Model
By
–
I led the creation of the very first high-resolution AI-based weather model FourCastNet at @NVIDIA @Caltech in 2021. Instead of bottoms-up physics-based weather forecasting, for the first time, we were able to show that AI-based models are accurate and tens of thousands of times… pic.twitter.com/1Rp3RAfeln
— Prof. Anima Anandkumar (@AnimaAnandkumar) 21 juillet 2025I led the creation of the very first high-resolution AI-based weather model FourCastNet at @NVIDIA @Caltech in 2021. Instead of bottoms-up physics-based weather forecasting, for the first time, we were able to show that AI-based models are accurate and tens of thousands of times
-
Gemini Deep Think Wins Gold at International Mathematical Olympiad
By
–
Very excited to share that an advanced version of Gemini Deep Think is the first to have achieved gold-medal level in the International Mathematical Olympiad! , solving five out of six problems perfectly, as verified by the IMO organizers! It’s been a wild run to lead this
-
Building Transformers.js App with Kimi K2 and Groq
By
–
Kimi K2 + @GroqInc vibe coding in anycoder build a transformers.js app for Xenova/moondream2 model card info and example included in prompt app was deployed successfully on HF in one click
-
CSD-VAR: Content-Style Decomposition in Visual Autoregressive Models
By
–
CSD-VAR
— AK (@_akhaliq) 21 juillet 2025
Content-Style Decomposition in Visual Autoregressive Models pic.twitter.com/BByVwZq8zKCSD-VAR Content-Style Decomposition in Visual Autoregressive Models
-
Mono-InternVL-1.5: Efficient Monolithic Multimodal Language Models
By
–
Mono-InternVL-1.5 Towards Cheaper and Faster Monolithic Multimodal Large Language Models
-
Nested Matryoshka Clustering for Scalable Visual Representation Learning
By
–
Franca Nested Matryoshka Clustering for Scalable Visual Representation Learning