Framework for testing risks and biases in silicon sampling as it relates to disability and hard-to-reach populations in survey research.
Predict how a population will react. Simulates a crowd, polls every person, and tells you when the answer came from your prompt instead of the world.
Code, data and receipts for "A Validity Ladder for LLM-Simulated Populations": a gate and four levels that test each use of a persona-simulated sample against human data, with a worked example on 28,800 PHQ-8 assessments and three published silicon samples.
TruckMind is an autonomous AI operator for food truck startups. It uses census-grounded market research, synthetic customers, strategy simulation, dynamic pricing, and conversational ordering to turn a concept into a running business.
Preprint: Synthetic respondents in survey research — an empirical test of LLM 'digital twins' against a student survey (N = 1,795). Russian original + English version.
NVIDIA Nemotron Personas Korea 기반 소비자 조사 시뮬레이터 — 통계 정합 합성 페르소나를 선별해 설문·심층인터뷰를 대신 돌린다
To associate your repository with the silicon-sampling topic, visit your repo's landing page and select "manage topics."