1 paper
Kristina Šekrst, Virna Karlić
When users address large language models, they produce directive speech acts whose pragmatic features differ from those of both everyday conversation and traditional human-computer…