How Google Uses Large Language Models

The Google Tunix Hackathon on Kaggle challenged developers to transform small, non-reasoning base models into general reasoning engines using Kaggle TPUs and a limited compute budget. The winning teams achieved this by implementing multi-stage post-training pipelines that combined Supervised Fine-Tuning (SFT) with advanced alignment techniques like GRPO and SimPO. Ultimately, the competition democratized AI development by proving that highly capable, structured reasoning models can be successfully trained by the community using accessible, open-source resources.

Google has released version 1.0.0 of the Agent Development Kit (ADK) for Java, introducing powerful new features like Google Maps grounding, built-in URL fetching, and a standardized Agent2Agent protocol for cross-framework collaboration. The update enhances agent control through a new "App" and "Plugin" architecture, which allows for global logging, automated context window management via event compaction, and "Human-in-the-Loop" workflows for action confirmations. Additionally, the release provides robust session and memory services using Google Cloud integrations like Firestore and Vertex AI to manage long-term state and large data artifacts.

Google announces the public preview of the Developer Knowledge API and its associated Model Context Protocol (MCP) server, providing a canonical, machine-readable gateway to Google's official devel...

This tutorial demonstrates how to fine-tune FunctionGemma, a small language model for translating natural language into API calls, using Google's Tunix library on TPUs.

The article discusses the integration of the Apigee Operator for Kubernetes with the GKE Inference Gateway to enhance API management for AI and Large Language Models (LLMs).

The article discusses the integration of Google’s Agent Development Kit (ADK) for Java with the LangChain4j LLM framework, enabling developers to utilize a variety of Large Language Models (LLMs) f...

This article discusses the integration of Google's EmbeddingGemma model with Google Cloud's Dataflow to create a scalable embedding pipeline for AI applications.

This article provides a comprehensive guide on how to train a GPT-2 model using JAX on TPU, highlighting the ease of leveraging Google TPUs for free.

The article introduces GenAI Processors, an open-source Python library from Google DeepMind aimed at simplifying the development of sophisticated AI applications using Large Language Models (LLMs).

The article discusses Gemma, a family of lightweight generative AI models, and introduces the concept of Agentic AI, which allows AI to make proactive decisions and utilize external tools.

The article discusses the Vertex AI RAG Engine, a tool designed to help developers build grounded generative AI applications by addressing challenges like hallucinations and outdated knowledge.

This article discusses how to publish Keras models on Kaggle and Hugging Face, highlighting the ease of sharing fine-tuned models with the community.

Gemini 1. 5 Pro has been launched in over 180 countries, introducing features such as native audio understanding, system instructions, and JSON mode.

Season 3 of the People of AI podcast features a lineup of inspiring guests discussing their journeys in the field of Artificial Intelligence.

The article introduces MakerSuite, a tool designed to simplify the development of generative AI applications using Google's latest models.