Design Inspiration from kuleshov-group.github.io/blog/blog/2026/how-to-build-a-diffusion-language-model/

The Kuleshov Group's blog post 'How to Build a Diffusion Language Model' presents a technical guide on constructing diffusion-based language models, a class of generative AI that iteratively refines text. The page, hosted on kuleshov-group.github.io, uses Roboto and -apple-system fonts with a palette of #2698ba, #1c1c1d, and #e8e8e8, as captured by the design tool Fudge.

How to Build a Diffusion Language Model | Kuleshov Group A Fudge conversation about the captured design, fonts, and colors of How to Build a Diffusion Language Model | Kuleshov Group. Show me the design details captured from How to Build a Diffusion Language Model | Kuleshov Group. Captured pages This capture comes from kuleshov-group.github.io https://kuleshov-group.github.io/blog/blog/2026/how-to-build-a-diffusion-language-model/ . It uses Roboto and -apple-system . The captured palette contains 3 prominent colors. Colors 2698ba 1c1c1d e8e8e8 Fonts captured from kuleshov-group.github.io Roboto /share/font-roboto-2032 -apple-system Roboto How Fudge assembled this page Fudge automatically assembled this profile from the saved capture and its extracted typography and color observations. Unknown facts are omitted.