Deep Value Benchmark: Measuring Whether Models Generalize Deep values or Shallow Preferences

Distributive Fairness in Large Language Models: Evaluating Alignment with Human Values