Fig. 1 (IMAGE)
Caption
Schematic overview of the study design. The authors asked large language models (LLMs) to rate approximately 700 English words that are typically acquired early in childhood according to 21 psychological features (e.g., Concreteness, Emotional Valence). By comparing these AI-generated ratings with ratings from human participants in prior studies, the authors examined how well LLMs can reproduce humanlike impressions and intuition about these words.
Credit
Hiromichi Hagihara
Usage Restrictions
Credit must be given to the creator.
License
Original content