From the Lab

Field notes & experiments

A mix of articles and hands-on interactive pages built for (and by) this community.

Orange Labs interactive demo: how AI text watermarking works, word by word
Interactive

How AI Text Watermarking Works (Interactive Demo)

Claude now watermarks its text using Google DeepMind's SynthID-Text. Run the same tournament sampling mechanism yourself, word by word, and see how detection actually works.

Abstract visualization of training data
Article

What Are Large Language Models (LLMs)? [Part 2: Technical Aspects]

Learn about the technical aspects of LLMs and Transformer Architecture in Part 2 of our comprehensive guide. Explore the nuances of training datasets; their curation, ethical considerations, and their keystone role in language models.

A massive pile of books in a warehouse, representing the scale of text LLMs are trained on
Article

What Are Large Language Models (LLMs)? [Part 1]

Unlock the mysteries of LLMs like ChatGPT without any tech jargon! Dive into our easy-to-understand guide and lay the foundation for grasping these powerful AI language tools.