PrivacySIM shows that conditioning LLMs on user personas like demographics and attitudes improves simulation of privacy choices but reaches only 40.4% accuracy against real responses from 1,000 users.
Using large language models to simulate multiple humans and replicate human subject studies
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 2verdicts
UNVERDICTED 2roles
background 1polarities
background 1representative citing papers
Introduces a 1.2-million-character narrative dataset from 92 residents, benchmarks 18 LLMs on fidelity with life-history profiles, and presents curriculum-LoRA as a low-cost personalization method that matches high-fidelity baselines at 10x lower token cost.
citing papers explorer
-
PrivacySIM: Evaluating LLM Simulation of User Privacy Behavior
PrivacySIM shows that conditioning LLMs on user personas like demographics and attitudes improves simulation of privacy choices but reaches only 40.4% accuracy against real responses from 1,000 users.
-
Benchmarking LLMs for Community Governance Simulation with Life-history Narratives
Introduces a 1.2-million-character narrative dataset from 92 residents, benchmarks 18 LLMs on fidelity with life-history profiles, and presents curriculum-LoRA as a low-cost personalization method that matches high-fidelity baselines at 10x lower token cost.