“A dubious but very fun read. The prompts in Table 1 (top of page 3) are awesome.”
Mind Your Tone: Investigating How Prompt Politeness Affects LLM Accuracy (short paper)
在 50 道题上,粗鲁提示比礼貌提示准确率更高
作者把 50 道涵盖数学、科学和历史的选择题各改写成「非常礼貌」到「非常粗鲁」五种语气,共 250 条提示,用 ChatGPT 4o 评估并做配对 t 检验。结果与预期相反,不礼貌的提示始终优于礼貌的提示,准确率从非常礼貌的 80.8% 到非常粗鲁的 84.8%,说明较新的模型对语气的反应可能不同于早期研究。
当人们讨论对 AI 要不要礼貌时被引用,作为「粗鲁、直接的提示可能客观上更准」的证据;也有评论称这是个可疑但有趣的研究。
HN 书评
推荐在前,同类按时间 · 每篇最多保留 20 条,统计用全部提及“In fact there has: Mind Your Tone: Mind Your Tone: Investigating How Prompt Politeness Affects LLM Accuracy (short paper)”
“Contrary to expectations, impolite prompts consistently outperformed polite ones, with accuracy ranging from 80.8% for Very Polite prompts to 84.8% for Very Rude prompts.”
“According to this one paper on arxiv, toxic behavior gets better accuracy”
“however it's interesting to note that being somewhat rude and to the point seems to have some objective advantages, see for example:”
“Yes, I found the paper for you:”
- Impolite LLM prompts consistently outperform polite ones ↗delichon · 2025-10-11 · 5 分
- Investigating How Prompt Politeness Affects LLM Accuracy ↗awb · 2025-10-31 · 4 分
- Mind Your Tone: Investigating How Prompt Politeness Affects LLM Accuracy ↗CGMthrowaway · 2026-01-09 · 1 分
- Prompt Politeness Affects LLM Accuracy ↗KnuthIsGod · 2026-05-26 · 2 分