Scientists warn that current AI tests reward polite responses rather than real moral reasoning in large language models.
Some results have been hidden because they may be inaccessible to you
Show inaccessible resultsSome results have been hidden because they may be inaccessible to you
Show inaccessible results