Discover

Tools

Actions

How this feed is ranked

Candidates are drawn at random from curated Wikipedia categories, then scored with this model. Nothing is hand-picked and no metric is invented: every input is Wikipedia metadata.

Technical relevance25%: Technical vocabulary in the title, description, categories and intro.
Article quality20%: WikiProject assessment class and article length.
Conceptual depth20%: Density of formal concepts (theorems, models, algorithms) and curiosity cues.
Obscurity15%: Low 30-day pageviews, discounted when article quality is weak.
Topic diversity10%: Distance from topics already shown in this batch.
Exploration relevance10%: Strength of the link to the current topic or article.

Rejected outright: disambiguation pages, lists, stubs under 6k bytes, biographies, entertainment, sports, places and organisations. Pageviews cover the last 14 days.

General

Print/export

Printable version

In other projects

The internet’s deepest rabbit holes

A live, ranked stream of obscure and technically deep Wikipedia articles, sampled from computer science, mathematics, physics, engineering, biology and economics. Open one, and follow its links as far down as you like.

Topic:

Mathematics:

Rank by:

The rabbit hole is temporarily blocked.

Wikipedia is rate limiting us right now. Try again in a minute.

Topic:

Mathematics:

Rank by:

Discoveries

Discoveries

Discoveries