Computational social science
Empirically evaluating commonsense intelligence in large language models with large-scale human judgments