WorldValuesBench: A Large-Scale Benchmark Dataset for Multi-Cultural Value Awareness of Language Models

1 code implementation • 25 Apr 2024 • Wenlong Zhao, Debanjan Mondal, Niket Tandon, Danica Dillion, Kurt Gray, Yuling Gu

The awareness of multi-cultural human values is critical to the ability of language models (LMs) to generate safe and personalized responses.

Value prediction