Zheni Zeng

New methods improve adaptability of language models while retaining past knowledge.

Research aims to develop language models with unique personalities for better human-like interactions.

KBAlign helps machines learn faster and more effectively through self-questioning techniques.