Theme

The measurements were wrong, not the models

Explore this theme’s evidence over time →

Sources

Attribution and reporting linked by this record; links do not imply independent verification.

When an AI audit finds errors in 42% of a benchmark’s problems and every score jumps on correction, the apparent plateau was an artifact of the ruler. Capability that looked slow was already there.

Development timeline

Newest reporting dates first. Reporting dates may differ from event dates.

    Knowledge graphopen in browser →

    Cite or correct this record

    Suggest a correction — opens a GitHub draft for you to review and submit.