{"slug": "breaking-homogeneity-diversifying-persona-sets-for-creative-llm-outputs", "title": "Breaking Homogeneity: Diversifying Persona Sets for Creative LLM Outputs", "summary": "A new arXiv paper (2609.30492v1) reports that evolutionary persona generation increased response diversity by 78.8%, originality by 26.1%, flexibility by 49.5%, and holistic creativity by 13.9% over task-only prompting on the Alternative Uses Task, while maintaining 98.5% validity. The study formulates persona diversification as a set-level conditioning problem and compares selecting versus generating personas and space-filling versus frontier-seeking diversity across four methods. On Infinity-Chat, evolutionary personas nearly doubled persona-induced response separation relative to random personas and composed with creativity-optimized prompting to add 18.6% response diversity and 6.3% creativity.", "body_md": "arXiv:2609.30492v1 Announce Type: new \nAbstract: Language models often produce homogeneous responses to open-ended tasks; such homogeneity can spawn groupthink-the convergence of ideas toward a singular and potentially suboptimal decision. We formulate persona diversification as a set-level conditioning problem and study two orthogonal design choices: selecting versus generating personas, and space-filling versus frontier-seeking diversity. We instantiate this design space with four methods spanning coverage and dispersion subset selections, uniform-coverage sampling, and evolutionary persona generation. Evaluations on the Alternative Uses Task (AUT), Infinity-Chat, and Divergent Association Task (DAT) show the benefits of the proposed methods across tasks and creativity objectives. On AUT, evolutionary persona generation increases response diversity by 78.8%, originality by 26.1%, flexibility by 49.5%, and holistic creativity by 13.9% over task-only prompting, while maintaining 98.5% validity; on Infinity-Chat, it nearly doubles persona-induced response separation relative to random personas. Moreover, evolutionary personas compose with creativity-optimized prompting, further increasing its response diversity by 18.6% and creativity by 6.3%. These results establish persona-set geometry as a task-agnostic mechanism for eliciting divergent LLM outputs, and support persona diversification as a reusable complement to prompt optimization.", "url": "https://wpnews.pro/news/breaking-homogeneity-diversifying-persona-sets-for-creative-llm-outputs", "canonical_source": "https://arxiv.org/abs/2609.30492", "published_at": "2026-09-28 04:00:00+00:00", "updated_at": "2026-09-28 04:18:46.257175+00:00", "lang": "en", "topics": ["large-language-models", "ai-research", "generative-ai", "natural-language-processing"], "entities": ["arXiv", "Alternative Uses Task", "Infinity-Chat", "Divergent Association Task"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/breaking-homogeneity-diversifying-persona-sets-for-creative-llm-outputs", "markdown": "https://wpnews.pro/news/breaking-homogeneity-diversifying-persona-sets-for-creative-llm-outputs.md", "text": "https://wpnews.pro/news/breaking-homogeneity-diversifying-persona-sets-for-creative-llm-outputs.txt", "jsonld": "https://wpnews.pro/news/breaking-homogeneity-diversifying-persona-sets-for-creative-llm-outputs.jsonld"}}