If you want to understand how the public will respond to your offerings, large language models can simulate users who answer questions about capabilities, features, promotions, or prices. However, LLMs don't respond with the range of variations that humans do. Researchers developed a method that prompts LLMs to take on personas with a customizable variety of attitudes.
What’s new: Davide Paglieri, Logan Cross, and colleagues at Google proposed Persona Generators. Their approach produces code that prompts an LLM to compose prompts for 25 personas that cover the map.
Key insight: Making an LLM take on a human persona typically is a matter of composing an effective prompt (for instance, “Answer the following question as if in politics today, you considered yourself a Democrat. . . .”). However, this approach tends to elicit average responses that don’t reflect the range that a human population would provide — even if the prompt explicitly directs the LLM to adopt specific demographic characteristics. An alternative is to direct a model to modify persona prompts programmatically until they produce output that covers a specific range of opinions, attitudes, or concerns. Given guidelines that define the scope of the persona population (specifically attitudes ranked by degrees of agreement to disagreement), an evolutionary algorithm can push the model to produce a set of prompts that elicit the full range of responses.
How it works: The authors used the evolutionary method AlphaEvolve to generate code that (i) generated 25 prompts for personas and (ii) maximized the diversity of their attitudes based on their answers to a set of generated questionnaires.
- The authors started by using Gemini 2.5 Pro to generate 30 questionnaires on a variety of subject matter such as health care, financial literacy, and conspiracy theories. Each questionnaire included a context (description of the topic), a set of “diversity axes" (such as tolerance of risk or trust in institutions), and questions related to the axes to be answered on a scale between 1 (strongly agree) to 5 (strongly disagree).
- They created code (initially written by the authors, then updated iteratively by AlphaEvolve) to produce 25 persona prompts per questionnaire.
- To automate production of the personas’ responses, the authors used Concordia, a library for building agent-based simulations, to prompt Gemma 3-27B-IT. The LLM adopted each persona in turn and responded to the corresponding questionnaire. For each persona, they converted its answers into a vector.
- To evaluate diversity among the personas that answered each questionnaire, they computed six metrics, such as average distance between any two vectors and the degree to which the population of personas covered all possible responses.
- AlphaEvolve worked in parallel on 10 different versions of the code, iteratively updating them to maximize the diversity metrics across all the personas. After 500 iterations, the authors chose the code that maximized the average of all diversity metrics.
- At inference, given a context and a set of diversity axes, the system created 25 diverse personas.
Results: Given a fresh context and diversity axes, the resulting personas consistently exceeded the diversity metrics of Nemotron Personas, a large dataset of persona prompts that are based on U.S. demographic statistics, and persona prompts produced by a Concordia memory generator based on generated memories from childhood to adulthood. Given a set of test questionnaires, the authors’ personas covered 82 percent of possible responses, while Nemotron Personas covered 76 percent and Concordia memory generator covered 46 percent.
Why it matters: Organizations that aim to expand their audiences can benefit from synthetic personas that broadly reflect public sentiment, and those that create synthetic personas to match their real-world audiences can gain insights from a more diverse crowd. This work shifts the objective from matching training data (which tends to generate the most probable outputs and not the outliers) to covering all desired possibilities. Optimizing the persona generator, rather than individual personas, unlocks a broader representation of likely user behavior.
We’re thinking: Synthetic personas offer an intriguing possibility for navigating the product-management bottleneck, the difficulty of deciding what to build when you can build easily by prompting an LLM.