4 ms·Large Language Models Fail on Trivial Alterations to Theory-of-Mind Tasks3 points by haltist 3y ago