原帖引用 · AI 翻译
标题：“大型语言模型会对道德价值观排序吗？一项道德能力的传递性检验”
这篇论文发现，大型语言模型的道德选择往往缺乏稳定的价值观排序，而将其回答取平均值会制造出一种虚假的、类似人类的道德层级。
许多研究认为，大型语言模型在道德测试中表现良好，因此具有合理的道德价值观。但给出好的答案并不能说明模型会如何权衡相互冲突的价值观。
研究人员向包括 GPT-5.2、Gemini 和 DeepSeek 在内的 5 个大型语言模型提供成对的道德违规行为，并询问哪一种更严重。他们直接从模型的下一词元输出中读取每个答案的概率，然后检验这些选择是否符合一致的排序。
颠倒 2 个选项的顺序后，30 个测试组中有 26 个的排序发生了变化。取平均值后，5 个模型中有 4 个呈现出清晰、几乎相同的排序，但这种排序与任何单独测试集的结果都不一致。
在信任模型的道德对齐评分之前，应检查它在单条提示和两种答案顺序下的选择是否保持一致的排序，且不要取平均值。
原文
This paper finds that LLM moral choices often lack a stable ranking of values, and averaging their answers creates a fake, human-like moral hierarchy. Many studies conclude that LLMs hold sensible moral values because they do well on morality tests. But good answers do not show that a model weighs competing values against each other. The researchers gave 5 LLMs, including GPT-5.2, Gemini, and DeepSeek, pairs of moral violations and asked which was worse. They read each answer's probability straight from the model's next-token output, then tested whether the choices fit a consistent ranking. Reversing the order of the 2 options changed the ranking in 26 of 30 test blocks. After averaging, 4 of 5 models showed a clean, nearly identical ranking that matched no individual test set. Before trusting a model's moral alignment scores, check that its choices hold a consistent ranking on individual prompts and both answer orders, without averaging.