Junjie Hu

Study reveals language models can generate useful PET report impressions.

A new method enhances learning from unlabeled data in diverse domains.

Innovative model editing techniques offer safer language model outputs without extensive retraining.

Examining how language models recognize cultural similarities across diverse societies.

Assessing the role of multilingual models in supporting bilingual students.

New mining tactics threaten fairness in cryptocurrency mining.

A new method aligns language models with diverse group preferences.