Illuminating the Black Box: A Psychometric Investigation into the Multifaceted Nature of Large Language Models