分解语言模型中的事实性谄媚：规模与指令调优如何塑造鲁棒性
Decomposing Factual Sycophancy in Language Models: How Size and Instruction Tuning Shape Robustness
AI总结：
通过将事实性谄媚分解为真值边际和操纵敏感性两个通道，研究了模型规模和指令调优对56个开源语言模型（0.3B-32B参数）在13种操纵类型下鲁棒性的影响。
摘要：
事实性谄媚是指语言模型在社会压力下放弃正确、可验证答案的现象。由于只有当朝向错误答案的压力超过模型对真相的中立偏好时才会发生翻转，翻转率混淆了两种机制：基线偏好强度（真值边际）以及压力将其偏移的程度（操纵敏感性）。我们将事实性谄媚分解为这些通道，并用它们来分离规模和指令调优对56个开源权重模型（参数范围0.3B-32B，13种操纵类型）的影响。我们发现脆弱性主要由规模决定，但指令调优改变了规模的作用方式：小的指令调优模型可能变得不那么鲁棒，而大的指令调优模型通常变得更鲁棒。指令调优主要增加真值边际，但其行为效果取决于操纵类型。缩放对两个通道的影响也不同：基础模型获得边际但变得略微更易受操纵影响，而指令调优模型更快地获得边际并变得不那么敏感。因此，事实性谄媚不是一个单一的标量属性。评估应报告通道特定、操纵特定和规模条件下的鲁棒性，而不仅仅是翻转率。
英文摘要：
Factual sycophancy occurs when a language model abandons a correct, verifiable answer under social pressure. Because a flip occurs only when pressure toward a false answer exceeds the model's neutral preference for the truth, flip rates conflate two mechanisms: the strength of that baseline preference (truth margin), and how far pressure shifts it (manipulation sensitivity). We decompose factual sycophancy into these channels and use them to separate the effects of size and instruction tuning across 56 open-weight models spanning 0.3B-32B parameters and 13 manipulation types. We find that vulnerability is governed mainly by size, but instruction tuning changes how size acts: small instruction-tuned models can become less robust, whereas large instruction-tuned models usually become more robust. Instruction tuning primarily increases truth margin, but its behavioral effect depends on manipulation type. Scaling also changes the two channels differently: base models gain margin but become mildly more manipulation-sensitive, whereas instruction-tuned models gain margin faster and become less sensitive. Factual sycophancy is therefore not a single scalar property. Evaluations should report channel-specific, manipulation-specific, and size-conditioned robustness rather than flip rates alone.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
