AnalysisAI ModelsNovember 16, 2026

Nvidia's Nemotron-Labs-Sigma 8B is a diffusion language model

Read original source →reddit.com

Nemotron-Labs-Sigma-8B-instruct is an 8-billion-parameter continuous diffusion language model that denoises Gaussian noise in a 16-D token-embedding space instead of generating one token at a time. A 3B variant is also referenced.

1 source

More stories today

Open the live feed