Hi, I'm Jens!

University of Mannheim

L 15, 1–6

68161 Mannheim

I am a PhD student at the Chair for Data Science in the Economic and Social Sciences at the University of Mannheim. Our chair focuses on developing methods for understanding human social and economic behavior via analyzing textual and relational data.

My academic background bridges two distinct but complementary fields, with a B.A. in Sociology from Otto-Friedrich University Bamberg and an M.Sc. in Data Science from the University of Mannheim. This combination drives my research in Computational Social Science, where I focus on the deployment and evaluation of Large Language Models. Specifically, I am interested in two key areas: leveraging and evaluating LLMs for synthetic data generation to model society, and advancing knowledge in survey methodology. I pursue computational, empirical and data-driven research approaches in collaboration with colleagues from the computer sciences, the economic sciences and the social sciences.

news

Jul 28, 2026	I am co-organizing a tutorial at the 12th International Conference on Computational Social Science (IC²S²) in Burlington, Vermont. Tutorial: An Introduction to Simulating Human Survey Responses with Large Language Models: Potentials and Pitfalls This hands-on session introduces researchers to employing large language models for generating survey data with an emphasis on methodological discipline. We address how “silicon samples” can supplement human data collection through pretesting and statistical correction. Topics covered include: Systematic exploration of how design decisions affect simulation results The QSTN framework for structured simulations Python-based practical exercises for generating simulated responses Use cases including missing-data imputation and statistical correction Critical assessment of limitations, validation issues, and ethical considerations The tutorial is co-organized together with Georg Ahnert, Maximilian Kreutner, Indira Sen, Markus Strohmaier (University of Mannheim) and Kristina Gligorić (Johns Hopkins University).
Jul 03, 2026	I will present our paper Prompt Perturbations Reveal Human-Like Biases in Large Language Model Survey Responses at the Seventh Workshop on NLP and Computational Social Science (NLP+CSS) at ACL 2026 in San Diego. We test 18 large language models on survey questions from the World Values Survey, applying ten different perturbations to question wording and answer options across over 334,800 simulated survey interviews. We find that 17 out of 18 models display recency bias, disproportionately favoring the last-presented answer option, while larger models show greater overall robustness. Our results highlight the need for careful prompt design when generating synthetic survey data with LLMs.
Jun 15, 2026	I attended the CSS School at Lake Como and presented our work on how the level of human response variation in a survey items affects the alignment of synthetic survey responses with persona prompting.
May 20, 2026	I attended the CSS DACH Conference in Vienna, Austria, bringing together researchers in Computational Social Science from German-speaking countries.
May 11, 2026	I presented our paper at LREC 2026 in Palma de Mallorca, Spain: German General Social Survey Personas: A Survey-Derived Persona Prompt Collection for Population-Aligned LLM Studies

selected publications

German General Social Survey Personas: A Survey-Derived Persona Prompt Collection for Population-Aligned LLM Studies

Jens Rupprecht, Leon Fröhling, Claudia Wagner, and 1 more author

Nov 2025

Accepted at LREC 2026

Abs DOI PDF

The use of Large Language Models (LLMs) for simulating human perspectives via persona prompting is gaining traction in computational social science. However, well-curated, empirically grounded persona collections remain scarce, limiting the accuracy and representativeness of such simulations. Here we introduce the German General Social Survey Personas (GGSS Personas) collection, a comprehensive and representative persona prompt collection built from the German General Social Survey (ALLBUS). The GGSS Personas and its persona prompts are designed to be easily plugged into prompts for all types of LLMs and tasks, steering models to generate responses aligned with the underlying German population. We evaluate GGSS Personas by prompting various LLMs to simulate survey response distributions across diverse topics, demonstrating that GGSS-guided LLMs outperform state-of-the-art classifiers, particularly under data scarcity. Furthermore, we analyze how the representativity and attribute selection within persona prompts affect alignment with population responses. Our findings suggest that GGSS Personas provides a potentially valuable resource for research on LLM-based social simulations that enables more systematic explorations of population-aligned persona prompting in NLP and social science research.
Prompt Perturbations Reveal Human-Like Biases in LLM Survey Responses

Jens Rupprecht, Georg Ahnert, and Markus Strohmaier

Jul 2025

Abs DOI PDF

Large Language Models (LLMs) are increasingly used as proxies for human subjects in social science surveys, but their reliability and susceptibility to known response biases are poorly understood. This paper investigates the response robustness of LLMs in normative survey contexts – we test nine diverse LLMs on questions from the World Values Survey (WVS), applying a comprehensive set of 11 perturbations to both question phrasing and answer option structure, resulting in over 167,000 simulated interviews. In doing so, we not only reveal LLMs’ vulnerabilities to perturbations but also reveal that all tested models exhibit a consistent }textit{recency bias} varying in intensity, disproportionately favoring the last-presented answer option. While larger models are generally more robust, all models remain sensitive to semantic variations like paraphrasing and to combined perturbations. By applying a set of perturbations, we reveal that LLMs partially align with survey response biases identified in humans. This underscores the critical importance of prompt design and robustness testing when using LLMs to generate synthetic survey data.
QSTN: A Modular Framework for Robust Questionnaire Inference with Large Language Models

Maximilian Kreutner, Jens Rupprecht, Georg Ahnert, and 2 more authors

Dec 2025

Accepted at EACL 2026 System Demonstrations

Abs DOI

We introduce QSTN, an open-source Python framework for systematically generating responses from questionnaire-style prompts to support in-silico surveys and annotation tasks with large language models (LLMs). QSTN enables robust evaluation of questionnaire presentation, prompt perturbations, and response generation methods. Our extensive evaluation (>40 million survey responses) shows that question structure and response generation methods have a significant impact on the alignment of generated survey responses with human answers. We also find that answers can be obtained for a fraction of the compute cost, by changing the presentation method. In addition, we offer a no-code user interface that allows researchers to set up robust experiments with LLMs without coding knowledge. We hope that QSTN will support the reproducibility and reliability of LLM-based research in the future.
"Charging Process Chain"

Felix Röckle, Marco Raul Soares Amorim, Lukas Keicher, and 2 more authors

Dec 2022

Abs DOI PDF

As part of the "Wirkkette Laden" funding project, a holistic approach for analysing weak points in the charging ecosystem was developed and applied in collaboration between research and industry. The approach is based on four building blocks: Statistical data analysis, a diary study, a charging experiment, and an AI-based analysis of online user comments. The results show that there are challenges both in terms of the individual components of the charging ecosystem, such as charging stations, and in terms of the interfaces between the components within the system.