AI models challenge humans in understanding minds, but struggle with subtleties, study finds

A study in Nature Human Behaviour compares the theory of mind capabilities of GPT-3.5, GPT-4, and LLaMA2-70B against humans, finding that AI models show varying degrees of success across different tasks.

Related Keywords

Helsinki ,Eteläuomen Läi ,Finland , ,Nature Human Behaviour ,Helsinki Declaration ,Large Language Model Meta ,Holm Corrected Wilcoxon ,