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 ...

Google boosts Gemini 3 Deep Think AI and it’s a huge milestone for 3D printing

Gemini 3 Deep Think is focused on scientific and engineering work, and it's now now available to Google AI Ultra subscribers in the Gemini app. The post Google boosts Gemini 3 Deep Think AI and it’s a ...
The Lancet

Adaptive deep brain stimulation in Parkinson's disease

With the introduction of adaptive deep brain stimulation (aDBS) for Parkinson's disease, new questions emerge regarding who, why, and how to treat. This paper outlines the pathophysiological rationale ...