Ask Again, Then Fail: Large Language Models' Vacillations in Judgement

We observe that current conversational language models often waver in their judgements when faced with follow-up questions, even if the original judgement was correct. This wavering presents a significant challenge for generating reliable responses and building user trust. To comprehensively assess this issue, we introduce a \textsc{Follow-up Questioning Mechanism} along with two metrics to quantify this inconsistency, confirming its widespread presence in current language models. To mitigate this issue, we explore various prompting strategies for closed-source models; moreover, we develop a training-based framework \textsc{Unwavering-FQ} that teaches language models to maintain their originally correct judgements through synthesized high-quality preference data. Our experimental results confirm the effectiveness of our framework and its ability to enhance the general capabilities of models.