Guide Labs debuts a new kind of interpretable LLM

The company open-sourced an 8 billion parameter LLM, Steerling-8B, trained with a new architecture designed to make its ...

Researchers baked 3x inference speedups directly into LLM weights — without speculative decoding

Researchers from the University of Maryland, Lawrence Livermore, Columbia and TogetherAI have developed a training technique that triples LLM inference speed without auxiliary models or infrastructure ...
EDN

A real-world approach for AI-driven semiconductor manufacturing

Combine scalable analytics with advanced AI capabilities like LLMs and agentic tasks to create a new chipmaking platform.