engblogs

This blogpost discusses the challenge of out-of-distribution blindness in machine learning models and presents an energy-based approach for improving out-of-distribution detection.

Redesigning Snorkel's interactive machine learning systems to meet customer needs and enable the adoption of foundation models and large language models.

Overview of EleutherAI's approach to alignment research

Using physics and machine learning, researchers create an AI-based estimator to analyze rough particle surfaces in pharmaceutical pills and powders for manufacturing medicine.

Explanation of how arm64 infrastructure was bootstrapped using Zig cc toolchain

Exploring ways to enhance revenue generation, market access, and operational efficiency through a global payments suite, platform optimization, and developer tools.

A detailed analysis of persistent malware campaigns targeting businesses, including NodeStealer, DuckTail, and newer malware posing as ChatGPT, and the strategies to defend against them.

AI's compute fragmentation and the lessons learned from matrix multiplication

An invitation to provide feedback and ideas for the Modular framework

Exploring the crucial role of modularity in advancing artificial intelligence.

An update on the lock icon in Chrome's address bar, discussing the decision to replace it with a new 'tune' icon to better emphasize security as the default state and improve accessibility to site settings.

Learn how to build custom experiences and run a custom POS using the all-in-one Stripe Reader S700 device.