Checkpoint
Moral result
Across yes/no, number, percent, scale, years, and never answers.
Your answer vs. the closest AI model on each question