Qiongkai Xu

A novel approach to ensure privacy while maintaining text utility in NLP models.

Examining vulnerabilities in EaaS and new watermarking defense strategies.

A new task focuses on identifying machine-written sentences in human-AI mixed texts.