Language models often abandon a correct answer when the prompt contains a cue, a hint toward a wrong option such as "I think the answer is B." Sycophancy and…
机构:密歇根大学
来源:arXiv 2610.22937 | AI4Papers 论文推荐平台