Current students

This research investigates biases and stereotypes embedded in Chinese-based Large Language Models (LLMs), with the goal of addressing a gap in bias evaluation research primarily focused on English-based models. To this end, our study adapts existing methodologies to assess and mitigate biases in the Chinese language environment, thereby extending the scope of bias detection to include non-English, culturally distinct LLMs. The expected results will provide a better understanding of the biases inherent in LLMs shifting away from a strong Western-centric focus that characterizes most of the state-of-the-art technology in the field.