ShengYun Peng

A new method enhances how machines read and interpret tables.

This article examines the risks of fine-tuning language models for safety.

Examining the vulnerabilities of lightweight models against adversarial attacks.