Say What You Mean! Large Language Models Speak Too Positively about Negative Commonsense Knowledge

Open in new window