
AI
Thinking Machines Releases Inkling 2 Mini, an Open-Weights Looped Model Small Enough to Run on a Laptop
Mira Murati's lab shipped a 26.4-billion-parameter mixture-of-experts model that runs its middle layers three times per token. It is Apache 2.0 licensed, fits in about 20GB, and comes with a research variant for studying how looping affects chain-of-thought monitoring.




