Google DeepMind
4 reportsThe research story of Google DeepMind is traced through computing infrastructure; deployment and safety practices; and research publications. The treatment of research publications keeps independent testing separate from technical reports until their assumptions and scales are compared; the evidence cannot remove the concern that company demonstrations and marketing claims may not reflect performance in broader settings.
AI Decoder Surpasses PyMatching on Google Quantum Surface-Code Data
Quantum X Labs has reported that its AI-based quantum error correction decoder achieved higher decoding accuracy than standard matching algorithms on Google's public surface-code dataset, using only synthetic training and GPU acceleration
Google Deploys SL2T Model for Real-Time Sign Language Translation
Google has introduced SL2T, a sign-language-to-text model trained on over 100,000 hours of data, enabling American Sign Language users to translate gestures into English text on Pixel 11 devices using Gboard and Live Transcribe
Google DeepMind Expands Gemini Robotics to Full-Body Humanoid Control
Google DeepMind has released Gemini Robotics 2, a model designed to control entire humanoid robots, enabling walking, object manipulation, and multi-robot coordination with limited retraining across different robot platforms
AI Systems Challenge Human Role in Mathematical Discovery
Recent advances in large language models and proof assistants have enabled AI systems to generate, formalize, and verify complex mathematical proofs, raising new questions about the future of human mathematicians and the value of human understanding in mathematics