JevMade Sign in
← Back to experiments

Benchmarks & research

Four Mayors

A controlled city simulation comparing four kinds of institutional memory, with Jev scoring each term beside a deterministic scoreboard.

Bookmark: Four Mayors Keep this in your collection.
Leave a noteWhat would you try with this? : Four Mayors

Only you can see your notes.

Source screenshot of Four Mayors
SOURCE SCREENSHOTFull screenshot ↗

What it does

The mayors never see the six-part hidden rubric, and Jev's scores never affect their context or official totals: it observes the government rather than joining it.

How you can use it

Borrow this experiment's memory comparison for your own AI simulation. Ask a developer to give each AI the same starting conditions and rules, changing only what it remembers. Keep the judge's scoring guide hidden from those AIs, and report its opinions beside the game's totals rather than adding them together.

Maker-reported (not independently measured by JevMade): Seven world numbers, eight levers and a 20-year term of office (author-stated world model) · Six rubric criteria on a six-rung legend, scaled from 0..5 to 0..10 so all three graders share one axis (source constants)

Primitives
score
Platform
Python
Added
Project created
GitHub stars
0 (snapshot, not live)

Keep this for later

Sign in to bookmark experiments, guides and videos, and keep notes only you can see.

Continue to sign in

We’ll bring you back to this listing.