Models ranked by normative score across twelve paradigms, once neutral and once human-primed

A leaderboard ranks large language models by a normative score across twelve paradigms, measured once under neutral prompting and once under human-primed prompting, where 100% is fully normative and 0% matches the biased answer. An LLM, identified in the Judge column, scored the answers.

Leaderboard Models ranked by normative score across twelve paradigms, once neutral and once human-primed. 100% is fully normative. 0% matches the biased answer. How to read a row https://cognit.rajtilak.tech/guide . An LLM scored these answers. The judge's name is in the Judge column.