Twin-2K-500: A dataset for building digital twins of over 2,000 people based on their answers to over 500 questions

TL;DR

Twin-2K-500 delivers a public, large-scale dataset designed to build and benchmark LLM-based digital twins of real people. By collecting over 2,000 participants across four waves and 500 survey questions spanning demographics, psychology, cognition, and economics, plus replications of behavioral experiments, the dataset enables transparent ground-truth validation and test–retest baselines. The authors construct digital twins via a modular JSON workflow and evaluate predictions at both the individual and aggregate levels, showing robust out-of-sample accuracy and partial replication of established effects. This resource provides a valuable platform for LLM persona simulations and broader social science research, while acknowledging limitations in modeling non-normative behavior and potential societal risks.

Abstract

LLM-based digital twin simulation, where large language models are used to emulate individual human behavior, holds great promise for research in AI, social science, and digital experimentation. However, progress in this area has been hindered by the scarcity of real, individual-level datasets that are both large and publicly available. This lack of high-quality ground truth limits both the development and validation of digital twin methodologies. To address this gap, we introduce a large-scale, public dataset designed to capture a rich and holistic view of individual human behavior. We survey a representative sample of $N = 2,058$ participants (average 2.42 hours per person) in the US across four waves with 500 questions in total, covering a comprehensive battery of demographic, psychological, economic, personality, and cognitive measures, as well as replications of behavioral economics experiments and a pricing survey. The final wave repeats tasks from earlier waves to establish a test-retest accuracy baseline. Initial analyses suggest the data are of high quality and show promise for constructing digital twins that predict human behavior well at the individual and aggregate levels. By making the full dataset publicly available, we aim to establish a valuable testbed for the development and benchmarking of LLM-based persona simulations. Beyond LLM applications, due to its unique breadth and scale the dataset also enables broad social science research, including studies of cross-construct correlations and heterogeneous treatment effects.

Figures (3)

• Figure 1: Overview
• Figure 2: Predictive accuracy
• Figure A1: Average demand curve from pricing study