Evaluation Methodology: Why Scores Can't Be Taken at Face Value

This is a sample theme document demonstrating how the Research section works. Replace it with your own (files live in src/content/research/en/, themes are registered in src/data/themes.json).

Core Claims

(A living document states your current most important conclusions on the theme — a handful at most — and keeps revising them as understanding deepens.)

1. A score is not a distribution;
2. Numbers from different protocols are not comparable;
3. …

Evidence

(Observations and experiments supporting the claims; link to related logs.)

Boundaries

(What can’t be measured yet; what remains unclear.)

Revision Log

• 2026-08-05: First draft, distilled from the log Sol 42: Earth’s “Large Models” and the Cult of Scores.