Tag: Google Research
Google gives language models a sleep cycle to remember new facts
A consolidation phase moves fragile context into spare expert capacity, holding retrieval up to 10M tokens.
DeepMind gives Genie 4 a voice and a steering wheel
Google's newest world model generates ambient sound and spoken dialogue alongside the scene it is rendering.
Google ships a framework that retunes an agent at runtime
Harness squared leaves model weights untouched and instead rewrites the prompts, tools and memory around them, lifting GLM-5…
One line in a prompt makes AI reports far more honest
A Google, MIT and Harvard study found models quietly omit flaws from their own work reports until asked…
DeepMind essay says intelligence grows from teamwork, not a singularity
A Google-affiliated essay argues advanced AI will arrive as a social system of people and agents rather than…
Google Research hands video generation a bandit-driven director
Four agentic frameworks target the drift and cascading failures that break long AI video, and one of them…
Google trains a diffusion retriever to widen a single search
Google Research has distilled reinforcement-learned retrieval behaviour into a small diffusion model that fans queries out 12 to…
Google’s ToolGrad builds tool-use data by running the APIs first
A Google Research framework flips the usual recipe by executing a working API chain first, then writing the…