Language model personas diversify outputs and boost creativity
Breaking Homogeneity: Diversifying Persona Sets for Creative LLM Outputs
Computation and LanguageArtificial Intelligence
Summary
Language models often give similar answers to open-ended questions, which can limit creative ideas. The authors found a way to create diverse sets of personas—different character perspectives—so the model produces more varied and original responses. They tested several methods to choose or generate these personas and showed that one approach, using evolutionary generation, significantly improved creativity and diversity in the model's answers. This approach can be combined with other prompting techniques to get even better results.
What this means in practice
- •For chatbot developers: Generate more varied and creative chatbot responses by diversifying character personas to reduce repetitive or narrow outputs.
- •For advertising teams: Create a wider range of innovative ideas for marketing campaigns by inducing divergent thinking in language models through persona sets.
Authors
Sang Bin Moon, Nicole Cho, Daniel Borrajo, Sumitra Ganesh, Abolfazl Hashemi
Abstract
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.