
Projected Language Models: A Large Model Pre-Segmented Into Smaller Ones
Introducing Projected Language Models for Efficient Specialized Inference

PINE: Efficient Norm-Bound Verification for Secret-Shared Vectors
Efficient protocol PINE for norm-bound verification in secure aggregation of secret-shared vectors with low communication overhead

Creating and verifying stable AI-controlled systems in a rigorous and flexible way
Developing techniques for verifying stability in AI-controlled systems to enable safer deployment of robots and autonomous vehicles.

Defending sovereign workloads: Google Distributed Cloud's air-gapped approach to Zero Trust
Google Distributed Cloud's approach to Zero Trust for air-gapped environments in public sector and government organizations

Bringing cloud and AI capabilities to the tactical edge: Google Distributed Cloud air-gapped appliance is generally available
Introducing Google Distributed Cloud air-gapped appliance for cloud and AI at tactical edge environments.
28 portmanteau word blends (including a few surprises!)
Exploring the fascinating world of portmanteau word blends and their origins.

Modular: Debugging in Mojoš„
Explore powerful debugging features and tools in Mojo for improved development velocity.

Ferretv2: An Improved Baseline for Referring and Grounding
Enhancing referring and grounding with Ferretv2's advanced design features

AI method radically speeds predictions of materialsā thermal properties
AI method accelerates predictions of materialsā thermal properties for efficient power generation and microelectronics design.

On a Neural Implementation of Brenier's Polar Factorization
Implementing Brenier's Polar Factorization using Neural Networks for Machine Learning Applications

Accelerate your generative AI distributed training workloads with the NVIDIA NeMo Framework on Amazon EKS
Optimize generative AI distributed training workloads with the NVIDIA NeMo Framework on Amazon EKS

Cross-Cloud Network: Design global distributed applications at scale
Design global distributed applications at scale with Cross-Cloud Network