aiqbench analysis
Anamnesis represents a significant advancement in applying large language models to social science research by enabling the creation of demographically controllable virtual populations for survey simulation. By operationalizing structured narrative backstories, this open-source platform allows researchers to stress-test instruments and replicate real-world opinion distributions without relying on proprietary services or actual human subjects. This tool democratizes high-fidelity simulation capabilities, offering a transparent alternative for prototyping surveys in political science and other fields requiring complex demographic modeling.
Anamnesis: An Open-Source Platform for Large-Scale Backstory-Conditioned Survey Simulation
arXiv:2607.10628v2 Announce Type: replace-cross Abstract: We present Anamnesis, an interactive system for demographically controllable survey simulation using large language models. Open-source and designed for non-technical users/researchers, Anamnesis enables the prototyping and stress-testing of survey instruments on virtual populations rather than real human subjects. The platform operationalizes the recently introduced Anthology and Alterity frameworks, which use structured narrative backstories to condition model responses, within a unified web interface. It supports open-ended generation, probabilistic demographic resampling, and multimodal (image and audio) surveys. We evaluate the system through two case studies: (1) replicating segments of Pew Research Center's American Trends Panel (ATP) on political typology and biomedical issues and (2) emulating human preference in the New Yorker Caption Contest. In both cases, Anamnesis produces opinion distributions that more closely match real-world survey data than standard persona-prompting baselines, offering a transparent, reproducible, and open-source alternative to proprietary simulation services.
Abstract:We present Anamnesis, an interactive system for demographically controllable survey simulation using large language models. Open-source and designed for non-technical users/researchers, Anamnesis enables the prototyping and stress-testing of survey instruments on virtual populations rather than real human subjects. The platform operationalizes the recently introduced Anthology and Alterity frameworks, which use structured narrative backstories to condition model responses, within a unified web interface. It supports open-ended generation, probabilistic demographic resampling, and multimodal (image and audio) surveys. We evaluate the system through two case studies: (1) replicating segments of Pew Research Center's American Trends Panel (ATP) on political typology and biomedical issues and (2) emulating human preference in the New Yorker Caption Contest. In both cases, Anamnesis produces opinion distributions that more closely match real-world survey data than standard persona-prompting baselines, offering a transparent, reproducible, and open-source alternative to proprietary simulation services.
Accepted to EMNLP 2026 Demo. Platform site: this https URL
Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
arXiv:2607.10628 [cs.CL]
(or arXiv:2607.10628v2 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2607.10628 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both 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.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 249 words
