Researchers have developed Anamnesis, an open-source platform designed for simulating surveys on virtual populations using large language models. This system allows users to prototype and test survey instruments by conditioning model responses on structured narrative backstories, operationalizing the Anthology and Alterity frameworks. Anamnesis supports multimodal surveys and has demonstrated its effectiveness by replicating segments of Pew Research Center's American Trends Panel and emulating human preferences in the New Yorker Caption Contest, outperforming standard prompting baselines. AI
IMPACT This platform offers a novel, open-source method for researchers to test survey instruments on virtual populations, potentially reducing costs and improving survey design.
RANK_REASON The cluster contains an academic paper detailing a new open-source platform for survey simulation using LLMs. [lever_c_demoted from research: ic=1 ai=1.0]
- Alterity
- American Trends Panel Wave 42 - Trust in Science
- Anthology
- arXiv
- Hugging Face
- New Yorker Caption Contest
- Pew Research Center
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
