研究：LLM 道德选择缺乏稳定价值排序
一项论文发现，LLM 的道德选择往往缺乏稳定的价值排序，对答案取平均会造出虚假的、类人的道德层级。研究者让包括 GPT-5.2、Gemini、DeepSeek 在内的 5 个 LLM 对成对道德违规行为做比较，并直接从 next-token 输出读取概率；调换两个选项的顺序后，30 个测试块中有 26 个的排序发生变化。
Rohan Paul· @rohanpaul_ai · X·2026-11-11 12:08· 14 分钟前AI 评分47
AI 导读
一项论文发现，LLM 的道德选择往往缺乏稳定的价值排序，对答案取平均会造出虚假的、类人的道德层级。研究者让包括 GPT-5.2、Gemini、DeepSeek 在内的 5 个 LLM 对成对道德违规行为做比较，并直接从 next-token 输出读取概率。
正文 · 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.
来源：Rohan Paul · x.com
来源：RadarAI · aihot.news