1 paper
Chao Gao, Haijiang Liu, Qiyuan Li +3
Large language models often answer the same multiple-choice question inconsistently when it is posed under support-oriented and elimination-oriented framings. We investigate whethe…