About
About Marovi
Marovi is a multilingual technical wiki for machine learning, deep learning, and AI research. The goal is to make complex technical material understandable, accessible across languages, and continuously improving.
What you'll find
- Encyclopedia articles on foundational ML and AI concepts (gradient descent, attention mechanisms, dropout, transformers, etc.) — written for readers who already have some technical background and want a precise, well-cited explanation.
- Paper summaries of seminal and recent research papers, paired with the full source reproduction so you can read the technical detail alongside the editorial summary.
- A glossary of term-of-art definitions surfaced as hover-tooltips throughout the site. Hover any underlined term to see its short definition; click to read the full article if one exists.
- Multilingual content — every article and paper is available in English, Spanish, and Chinese, with carefully preserved terminology across the three languages.
Editorial approach
Marovi is AI-assisted but human-reviewed. AI tools draft articles, generate translations, and extract source reproductions from research papers. Editors review for accuracy, consistency, and terminology. Every page records its provenance in a `
` block so readers can see how a page was generated.
We aim for a precise, neutral, encyclopedic register — like Wikipedia but more focused on the ML/AI technical canon. We do not run advertising or sponsored content. We do not produce hot takes or news commentary; the goal is durable reference material.
Content license
All Marovi content is licensed under CC BY-SA 4.0. You may copy, redistribute, and adapt freely with attribution; see the Terms of Service for details.
Contributing
The site is open to contributions from approved editors. If you'd like to contribute (translation review, factual corrections, new articles), email hello@marovi.ai.
made by
Marovi is made by Felipe Felix Arias and a small group of contributors and AI assistants.