
How Skill-it! enables faster, better LLM training
How Skill-it! enables faster, better LLM training by optimizing data order and leveraging skills graph

Building Meta’s GenAI Infrastructure
Announcing Meta’s GenAI infrastructure with 24k GPU clusters for large-scale AI workloads and research advancements

Randomized Algorithms for Precise Measurement of Differentially-private, Personalized
Developing a new algorithm for precise and differentially-private personalized recommendations in the era of privacy-first systems.

Large language model inference over confidential data using AWS Nitro Enclaves
Protecting sensitive user data in large language model inference with AWS Nitro Enclaves

Run an audience overlap analysis in AWS Clean Rooms
Learn how to run secure audience overlap analysis using AWS Clean Rooms for media planning and partnership evaluation

Healthcare's AI transformation is already underway
Google Cloud advances healthcare AI tools with Vertex AI Search for Healthcare and Healthcare Data Engine, transforming patient care and optimizing data management.

Introducing Security Command Center Enterprise: The first multicloud risk management solution fusing AI-powered SecOps with cloud security
The first multicloud risk management solution merging AI-powered SecOps with cloud security

Improving ETAs with Multi-Task Models, Deep Learning, and Probabilistic Forecasts
Enhancing ETA accuracy with multi-task models, deep learning, and probabilistic forecasts.
Special report: What languages do the French study and why?
Exploring language learning trends among the French population using Duolingo data.

2024 BAIR Graduate Directory
Comprehensive guide to the latest trends and technologies in AI research

Announcing the General Availability of Databricks Feature Serving
Introducing Feature Serving for real-time AI applications with low latency and reduced operational overhead

Postgres Aurora DB major version upgrade with minimal downtime
Efficiently upgrade Postgres Aurora DB from v10 to v13 with minimal downtime using native logical replication and blue/green deploy strategy