The Hutter Prize
A compression contest based on the idea that better models of data make smaller files.
The task is to compress a Wikipedia corpus. The underlying argument is that doing this well requires finding more of the structure in the text.
Archived snapshot of prize.hutter1.net, October 18, 2026
I like being able to see the argument, dataset, rules, records, and writeups together. The constraints are part of the idea, and the old results are part of its history.
Source and preservation
- Source link
- Not checked
- Saved
- Details only
- On this page
- Link only
Topics: compression, ai, benchmarks, wikipedia, web-history
More from the Hall
OEIS — The On-Line Encyclopedia of Integer Sequences
An encyclopedia of integer sequences, with formulas, programs, comments, and cross-references.
Visit source ↗Internet Archive / Wayback Machine
Look up what a website said on an earlier date, or recover a page from a dead link.
Visit source ↗The Original Crichton Leprechaun News Story
A leprechaun in a tree, a local news crew, and perhaps the finest amateur sketch ever televised.
Watch surviving copy ↗