Xuandong Zhao

Investigating self-bias in LLMs and its impact on performance.

New watermarking methods improve text variety and detection in machine-generated content.

An overview of code hallucinations in LLMs and their impact on software development.