Conceptual structure coheres in human cognition but not in large language models

Open in new window