Eliciting Affective Events from Language Models by Multiple View Co-prompting