survey-simulation

Jev vs GPT-4.1 as synthetic survey respondents on Twin-2K-500. How you ask mattered more than which model you used.

한국어 가상 페르소나 설문 패널 — Nemotron-Personas-Korea 층화 표집 + Claude 응답 시뮬레이션 (가설 정제용)

LLM-powered questionnaire simulation toolkit with high-throughput CLI workflows (EasyPsych).

RSA-based fidelity map of demographic identity in an LLM (Mistral-7B): readable, faithful to Pew survey ground truth, and causally used are dissociable. arXiv:2608.18768