Pamela D. Rivière

Language models show a similar bias to humans in attributing false beliefs when knowledge is cued using specific verb types, suggesting this aspect of human reasoning may arise from statistical patterns in language.