Deep Learning
Image editing in Gemini just got a major improvement
Today in Gemini applicationIntroducing Google's recent DeepMind image editing model. People were driving bananas on this already in the early announcements - that is...
Deep Learning
Introducing Gemma 3 270M: a compact model that delivers incredibly powerful AI
The last few months have been an electrifying time for the Gemma family of open models. we presented Gemma 3...
Deep Learning
Rethinking how AI intelligence is measured
Current AI benchmarks struggle to keep pace with up-to-date models. While they are helpful in measuring model performance for specific tasks, it can be...
Deep Learning
Try Deep Think in the Gemini app
How Deep Think works: extending the parallel "thinking time" of GeminiJust as people solve sophisticated problems by taking the time to explore different points...
Deep Learning
Bringing artificial intelligence to the next generation of fusion energy
We are working with Commonwealth Fusion Systems (CFS) to bring the reality of spotless, safe and sound and unlimited fusion energy closer...
Deep Learning
Introducing CodeMender: an AI agent for code security
While vast language models are improving rapidly, code security bugs can be costly. CodeMender's automatic validation process ensures that code changes are...
Deep Learning
Gemma Scope 2: Helping the AI security community deepen its understanding of the behavior of elaborate language models
We are announcing a recent open source toolkit for interpreting language modelsVast language models (LLMs) are capable of incredible feats of reasoning,...
Deep Learning
Gemini Robotics 1.5 brings AI agents to the physical world
ThanksJean-Baptiste Alayrac, Months, Filemon Brakel, Anthony Brohan, Thomas Buschmann, Servant, Servant, B. The Ambrose, Southern Dasari, Todor Davchev, Abread Dave, Coline Devin,...