Google for Developers reposted this
From Foldables to Wearables: here's what's new from #GalaxyUnpacked! → https://goo.gle/4eUWUIh ⌚️ Get your app ready for the Galaxy Watch7 and Watch Ultra, running on Wear OS 5 📱 Build adaptive apps for a great user experience on the Galaxy Z Fold6 and Z Flip6 Build apps that look great across all these screen sizes and types! 🤩
Unlocking responsible AI with the new Gemma 2 family!→ https://goo.gle/3ywtgIV Today in the #Gemmaverse, we're releasing: 🛡 ShieldGemma: a suite of safety content classifiers for user protection 🔬 Gemma Scope: a new interpretability tool for transparency 👀 Gemma 2 (2B): a powerful, light-weight model with an updated 2 billion (2B) parameter size
🚀🤗 Download the new lightweight Gemma 2 (2B) model weights from Hugging Face. Explore the #Gemmaverse with new tools like ShieldGemma and Gemma Scope for safer and more interpretable AI → https://goo.gle/46C0eEg
Blockly developers, product managers, and learning designers met last month to talk about the future of coding education. Now available on-demand to all, watch the 20+ recorded sessions to catch up on what you missed! ➡️ https://lnkd.in/eeSWeRyK
Waking up and immediately sticking the landing on our coding tasks while still getting a top score. Anyone else? 🤸🏽♂️ #IYKYK
From a new #GeospatialCreator pilot in Google Maps to contributions to open standards like glTF, there were a lot of AR announcements at #GoogleIO. Have more questions? Drop them in the comments 👇
Any of these AR topics catch your eye? 👀 If you said yes, check out the latest updates for your topic of interest here → https://goo.gle/ar-io-24
🎉 Gemma 2 (2B) is small enough to run on edge devices! Learn more → https://goo.gle/4fvXVGU The new light-weight model offers: ✨ Cost-effective deployment ✨ Exceptional performance and accessibility ✨ Open and accessible entry point for exploring Gemma 2 capabilities
Gemma Scope: Look at the inner workings of Gemma 2 with sparse autoencoders. 🔧 → https://goo.gle/46wb6Ub Acting like a powerful microscope, Gemma Scope uses sparse autoencoders (SAEs) to zoom in on specific points within the model and make its inner workings more interpretable. Now devs can start building with: 🖥️ Transparency into the decision-making process of Gemma models 🔓 Open SAEs, with over 400 freely available, covering all layers of Gemma 2 (2B) and 9B 😌 An easy-to-use repository that offers code and examples for interfacing with SAEs and Gemma 2
Prepare for a whole lotta awesome. 🙌