研究人员发现大型语言模型（LLMs）存在严重的框架敏感性问题，事实相同但措辞不同的输入会导致LLMs改变其决策。引入了一个名为Fragile的新基准来测试这一点，结果显示在各种LLMs中平均决策翻转率为28.6%。简单的干预措施被证明无效，因此开发了Valign，一种针对内部模型表示以将决策锚定到稳定值并减轻框架效应的方法。 AI
影响 突显了LLMs在高风险决策中的一个关键漏洞，需要对鲁棒的对齐技术进行进一步研究。
排序理由 该集群包含一篇学术论文，详细介绍了LLM行为的新基准和缓解方法。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
完整方法见我们的编辑标准。
arXiv:2605.28188v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly deployed in high-stakes decision-making settings such as legal reasoning, where consistency under factually equivalent inputs is critical. However, we find that fact-preserved but differ…
Large Language Models (LLMs) are increasingly deployed in high-stakes decision-making settings such as legal reasoning, where consistency under factually equivalent inputs is critical. However, we find that fact-preserved but differently framed inputs can significantly destabiliz…