Mingzhe Li

Unsupervised Mitigating Gender Bias by Character Components: A Case Study of Chinese Word Embedding

2022

Word embeddings learned from massive text collections have demonstrated significant levels of discriminative biases. However, debiasing on the Chinese language, one of the most spoken languages, has been less explored. Meanwhile, existing literature relies on …

PAPILLON: Efficient and Stealthy Fuzz Testing-Powered Jailbreaks for LLMs

2024 · arXiv (Cornell University)

Large Language Models (LLMs) have excelled in various tasks but are still vulnerable to jailbreaking attacks, where attackers create jailbreak prompts to mislead the model to produce harmful or offensive content. Current jailbreak methods either …

VMSMO: Learning to Generate Multimodal Summary for Video-based News Articles

2020

A popular multimedia news format nowadays is providing users with a lively video and a corresponding news article, which is employed by influential news media including CNN, BBC, and social media including Twitter and Weibo. …