{"slug": "min-p-max-exaggeration-critical-analysis-of-min-p-sampling-in-language-models", "title": "Min-P, Max Exaggeration: Critical Analysis of Min-P Sampling in Language Models", "summary": "A re-examination of min-p sampling published on arXiv by Rylan Schaeffer and co-authors concludes that the evidence in Nguyen et al. 2024's \"Turning Up the Heat: Min-p Sampling for Creative and Coherent LLM Outputs\" fails to support claims that min-p improves quality, diversity, or a quality-diversity trade-off over basic, top-k, and top-p sampling. The reanalysis found the original human evaluations omitted data and ran statistical tests incorrectly, that sweeping the original NLP benchmarks showed min-p does not surpass baselines when controlling for hyperparameter count, that LLM-as-a-Judge evaluations lacked methodological clarity, and that adoption claims of 49k GitHub repositories and 1.1M GitHub stars were unsubstantiated and removed. The original authors ran a new human evaluation with a different implementation, task, and rubric that the reanalysis says still shows min-p does not improve over baselines, despite the paper's 18th-highest score at ICLR 2025 and Oral selection.", "body_md": "# Computer Science > Computation and Language\n\n  [Submitted on 16 Jun 2025 (\n\n[v1](https://arxiv.org/abs/2506.13681v1)), last revised 19 Jun 2025 (this version, v2)]\n# Title:Min-p, Max Exaggeration: A Critical Analysis of Min-p Sampling in Language Models\n\n[View PDF](https://arxiv.org/pdf/2506.13681)\n\n[HTML (experimental)](https://arxiv.org/html/2506.13681v2)\n\nAbstract:Sampling from language models impacts the quality and diversity of outputs, affecting both research and real-world applications. Recently, Nguyen et al. 2024's \"Turning Up the Heat: Min-p Sampling for Creative and Coherent LLM Outputs\" introduced a new sampler called min-p, claiming it achieves superior quality and diversity over established samplers such as basic, top-k, and top-p sampling. The significance of these claims was underscored by the paper's recognition as the 18th highest-scoring submission to ICLR 2025 and selection for an Oral presentation. This paper conducts a comprehensive re-examination of the evidence supporting min-p and reaches different conclusions from the original paper's four lines of evidence. First, the original paper's human evaluations omitted data, conducted statistical tests incorrectly, and described qualitative feedback inaccurately; our reanalysis demonstrates min-p did not outperform baselines in quality, diversity, or a trade-off between quality and diversity; in response to our findings, the authors of the original paper conducted a new human evaluation using a different implementation, task, and rubric that nevertheless provides further evidence min-p does not improve over baselines. Second, comprehensively sweeping the original paper's NLP benchmarks reveals min-p does not surpass baselines when controlling for the number of hyperparameters. Third, the original paper's LLM-as-a-Judge evaluations lack methodological clarity and appear inconsistently reported. Fourth, community adoption claims (49k GitHub repositories, 1.1M GitHub stars) were found to be unsubstantiated, leading to their removal; the revised adoption claim remains misleading. We conclude that evidence presented in the original paper fails to support claims that min-p improves quality, diversity, or a trade-off between quality and diversity.\n    \n\n## Submission history\n\nFrom: Rylan Schaeffer [\n[view email](https://arxiv.org/show-email/c6a816ae/2506.13681)]\n\n**Mon, 16 Jun 2025 16:38:04 UTC (933 KB)**\n\n[\\[v1\\]](https://arxiv.org/abs/2506.13681v1)\n**[v2]** Thu, 19 Jun 2025 04:41:52 UTC (939 KB)\n\n### Additional Features\n\n### References & Citations\n\nLoading...\n\n# Bibliographic and Citation Tools\n\nBibliographic Explorer \n\n*(*[What is the Explorer?](https://info.arxiv.org/labs/showcase.html#arxiv-bibliographic-explorer))\nConnected Papers \n\n*(*[What is Connected Papers?](https://www.connectedpapers.com/about))\nLitmaps \n\n*(*[What is Litmaps?](https://www.litmaps.co/))\nscite Smart Citations \n\n*(*[What are Smart Citations?](https://www.scite.ai/))\n# Code, Data and Media Associated with this Article\n\nalphaXiv \n\n*(*[What is alphaXiv?](https://alphaxiv.org/))\nCatalyzeX Code Finder for Papers \n\n*(*[What is CatalyzeX?](https://www.catalyzex.com))\nDagsHub \n\n*(*[What is DagsHub?](https://dagshub.com/))\nGotit.pub \n\n*(*[What is GotitPub?](http://gotit.pub/faq))\nHugging Face \n\n*(*[What is Huggingface?](https://huggingface.co/huggingface))\nScienceCast \n\n*(*[What is ScienceCast?](https://sciencecast.org/welcome))\n# Demos\n\n# Recommenders and Search Tools\n\nInfluence Flower \n\n*(*[What are Influence Flowers?](https://influencemap.cmlab.dev/))\nCORE Recommender \n\n*(*[What is CORE?](https://core.ac.uk/services/recommender))\n# arXivLabs: experimental projects with community collaborators\n\narXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.\n\nBoth individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.\n\nHave an idea for a project that will add value for arXiv's community? [**Learn more about arXivLabs**](https://info.arxiv.org/labs/index.html).", "url": "https://wpnews.pro/news/min-p-max-exaggeration-critical-analysis-of-min-p-sampling-in-language-models", "canonical_source": "https://arxiv.org/abs/2506.13681", "published_at": "2026-10-24 20:21:09+00:00", "updated_at": "2026-10-24 20:41:44.921302+00:00", "lang": "en", "topics": ["large-language-models", "ai-research", "machine-learning", "artificial-intelligence"], "entities": ["Rylan Schaeffer", "Nguyen et al.", "min-p", "ICLR 2025", "arXiv", "GitHub"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/min-p-max-exaggeration-critical-analysis-of-min-p-sampling-in-language-models", "markdown": "https://wpnews.pro/news/min-p-max-exaggeration-critical-analysis-of-min-p-sampling-in-language-models.md", "text": "https://wpnews.pro/news/min-p-max-exaggeration-critical-analysis-of-min-p-sampling-in-language-models.txt", "jsonld": "https://wpnews.pro/news/min-p-max-exaggeration-critical-analysis-of-min-p-sampling-in-language-models.jsonld"}}