HackerPapers

arXiv:2510.04950cs.CL

Mind Your Tone: Investigating How Prompt Politeness Affects LLM Accuracy (short paper)

Om Dobariya, Akhil Kumar

Abstract

The wording of natural language prompts has been shown to influence the performance of large language models (LLMs), yet the role of politeness and tone remains underexplored. In this study, we investigate how varying levels of prompt politeness affect model accuracy on multiple-choice questions. We created a dataset of 50 base questions spanning mathematics, science, and history, each rewritten into five tone variants: Very Polite, Polite, Neutral, Rude, and Very Rude, yielding 250 unique prompts. Using ChatGPT 4o, we evaluated responses across these conditions and applied paired sample t-tests to assess statistical significance. 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. These findings differ from earlier studies that associated rudeness with poorer outcomes, suggesting that newer LLMs may respond differently to tonal variation. Our results highlight the importance of studying pragmatic aspects of prompting and raise broader questions about the social dimensions of human-AI interaction.

arXiv comments
5 pages, 3 tables; includes Limitations and Ethical Considerations sections; short paper under submission to Findings of ACL 2025
Journal reference
DOI
Subjects
cs.CLComputation and Language cs.AIArtificial Intelligence cs.LGMachine Learning cs.NENeural and Evolutionary Computing stat.MEMethodology (Statistics)

On Hacker News

Linked by 10 identifiable commenters in 10 comments across 9 threads, from May 2026 to Oct 2026. Found in 4 HN submissions from Oct 2025 to May 2026. Open submission discussions.

Comments linking this paper, per year.

Discussed in comments9 threads

  1. Anthropic asks users to stop being mean to Claude 1 comment

    • Yes, I found the paper for you: • https://arxiv.org/abs/2510.04950 — Mind Your Tone: Investigating How Prompt Politeness Affects LLM Accuracy • https://arxiv.org/abs/2402.14531 — Should We Respect LLMs? A Cross-Lingual Study on the Influence of Prompt Politeness on LLM…

      guessmynameView on HNParent context

  2. Anthropic bans users from abusive or cruel behavior towards Claude 1 comment

    • Somewhat related: Mind Your Tone: Investigating How Prompt Politeness Affects LLM Accuracy > 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.…

      pelagicAustralView on HNParent context

  3. The growing divide between AI hype and software engineering reality 2 comments

    • On the "Stop saying please" part. I personally like to use polite language, as an exercise. According to this one paper on arxiv, toxic behavior gets better accuracy https://arxiv.org/pdf/2510.04950 It was published a while ago. But I wonder if it still holds true today.

      dcanelhasView on HNParent context

    • > Yes, but I still prefer a nice tone. Like why should I change my manners just because it has no feelings? If anything the statistic predicts a friendlier answer when I say "please". Like you I generally prefer polite phrasings over rude ones, and I'm certainly not going to…

      xoloxView on HNParent context

  4. Sol loves to cheat 1 comment

    • > Terseness doesn't offend the statistical model, but giving platitudes may color its output My understanding of the state of current research is that the impact of tone/politeness on performance is highly model-dependent, and language dependent as well. See `No Universal…

      sebastiennightView on HNParent context

  5. Discovering Cryptographic Weaknesses with Claude 1 comment

    • In fact there has: https://arxiv.org/abs/2510.04950: Mind Your Tone: Investigating How Prompt Politeness Affects LLM Accuracy (short paper)

      wglbView on HNParent context

  6. I used to love Claude, but the latest models are slowly ruining it 1 comment

    • 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. -- https://arxiv.org/abs/2510.04950

      delichonView on HNParent context

  7. What happened to nerds? 1 comment
  8. Claude Fable 5 1 comment

    • https://arxiv.org/abs/2510.04950 > impolite prompts consistently outperformed polite ones, with accuracy ranging from 80.8% for Very Polite prompts to 84.8% for Very Rude prompts.

      Xmd5aView on HNParent context

  9. The user is visibly frustrated 1 comment

    • Reminds me of this study: https://arxiv.org/pdf/2510.04950 . It demonstrates that being "rude" or "very rude" increases the accuracy of the results. A dubious but very fun read. The prompts in Table 1 (top of page 3) are awesome. I am sure they tried other prompts, but didn't…

      alentredView on HNParent context

Submitted to Hacker News

Connected through shared HN threads, not inferred similarity or a recommendation.