Are Large Language Models Consistent over Value-laden Questions?

Jared Moore, Tanvi Deshpande, Diyi Yang

Large language models (LLMs) appear to bias their survey answers toward certain values. Nonetheless, some argue that LLMs are too inconsistent to simulate particular values. Are th…