Xinpeng Wang

~Xinpeng_Wang3

3

论文总数

1.5

年均投稿

6.3

平均评分

接收情况3/3

会议分布

NeurIPS

1

ICLR

1

COLM

1

发表论文 (3 篇)

20252 篇

Refusal Direction is Universal Across Safety-Aligned Languages

NeurIPS 2025Poster

Surgical, Cheap, and Flexible: Mitigating False Refusal in Language Models via Single Vector Ablation

ICLR 2025Poster

20241 篇

Look at the Text: Instruction-Tuned Language Models are More Robust Multiple Choice Selectors than You Think

COLM 2024Poster

合作者 (8)

Barbara Plank3 篇

Chengzhi Hu2 篇

Hinrich Schuetze1 篇

Mingyang Wang1 篇

Yihong Liu1 篇

Paul Rottger1 篇

Paul Röttger1 篇