Using cognitive models to reveal value trade-offs in language models

Open in new window