Stories by tosh

Copilot vs. Cody: Why context matters for code AI

Google RecurrentGemma: Moving Past Transformers [pdf]

Griffin: Mixing Gated Linear Recurrences with Local Attention

Kirby's Avalanche

RecurrentGemma: Mixture local attention and linear recurrences

Koboldcpp-1.62.1 adds support for Command-R+

Google CodeGemma: Open Code Models Based on Gemma [pdf]

Embedding models

X Rolls Out Passkeys Support to iPhone Users Worldwide

Please stop throwing money at AI founders with no commercial plan, besides AGI

Diffusion-RWKV: Scaling RWKV-Like Architectures for Diffusion Models

Apple Ferret-UI: Grounded Mobile UI Understanding with Multimodal LLMs

Vercel AI SDK 3.0 with Generative UI Support

Morphic: An AI-powered answer engine with a generative UI

llm.f90: LLM Inference in Fortran

Hello OLMo: A truly open LLM

Stable LM 2 12B

Amdahl's Law

StableLM-2-12B