Google CloudStanford launches platform for high-risk data on Google Cloud
Stanford launches Carina, a customizable high-risk data platform on Google Cloud, to reduce lead time for project setup and provide a secure on-premise compute environment for researchers.
DoorDashAddressing the Challenges of Sample Ratio Mismatch in A/B Testing
Addressing the Challenges of Sample Ratio Mismatch in A/B Testing: This blogpost explores the challenges of sample ratio mismatch in A/B testing, provides examples of failures, and discusses solutions to reduce SRM rate.
Google CloudC3D VMs with 4th Gen AMD EPYC Processors now GA, with 45 percent performance increase
C3D VMs with 4th Gen AMD EPYC Processors now GA, offering 45 percent performance increase
Dear Duolingo: What counts as a "universal" word? (the finale)
Exploring the concept of 'universal' words and analyzing reader suggestions to determine their universality.
MIT AINew technique helps robots pack objects into a tight space
New technique using generative AI solves packing problems for robots by satisfying multiple constraints simultaneously
GitHubMeasuring Git performance with OpenTelemetry
Enhance Git performance analysis using Trace2 receiver with OpenTelemetry for detailed telemetry visualization
Berkeley AIRethinking the Role of PPO in RLHF
Exploring new perspectives on the PPO algorithm in hierarchical reinforcement learning
DatabricksHow CIOs are laying the foundation for AI-led growth
CIOs are investing in and leveraging data and AI to lay the foundation for AI-led growth, with high expectations for efficiency gains.
Google AIImproving traffic evacuations: A case study
Improving traffic evacuations in cities using case study and traffic simulations
EleutherAILlemma: An Open Language Model For Mathematics
Llemma is an open language model for mathematics trained on a 55 billion token dataset that shows improved mathematical capabilities and can be adapted to various tasks through prompting or additional fine-tuning.
Berkeley AIRethinking the Role of PPO in RLHF
Rethinking the Role of PPO in RLHF: Introducing Pairwise Proximal Policy Optimization (P3O) to harmonize the reward learning and RL fine-tuning stages in Reinforcement Learning with Human Feedback (RLHF).
Google CloudGoogle Cloud (Apigee) named a Leader in the 2023 Gartner® Magic Quadrant™ for API Management
Google Cloud (Apigee) recognized as a Leader in the 2023 Gartner Magic Quadrant for API Management