Mingqian Feng

Ph.D. Candidate, University of Rochester CS

mingqian.feng@rochester.edu

Bio

I am a Ph.D. candidate in the Department of Computer Science at University of Rochester (UR), where I am fortunate to be advised by Prof. Chenliang Xu. Prior to that, I received my M.S.E degree in Financial Mathematics at Johns Hopkins University (JHU). and my B.Sc. degree in physics at University of Science and Technology of China (USTC).

My research focuses on AI Security, with an emphasis on training powerful red-teaming agents for rigorous stress testing and developing comprehensive evaluation frameworks for realistic risk assessment.

I am actively looking for collaboration on research. If you share similar interests, please feel free to contact me!

News

Publications

Google Scholar.
* indicates equal contribution.

Statistical Estimation of Adversarial Risk in Large Language Models under Best-of-N Sampling

Statistical Estimation of Adversarial Risk in Large Language Models under Best-of-N Sampling

Mingqian Feng, Xiaodong Liu, Weiwei Yang, Chenliang Xu, Christopher White, Jianfeng Gao

NeurIPS 2026.

Paper
MultiBreak: A Scalable and Diverse Multi-turn Jailbreak Benchmark for Evaluating LLM Safety

MultiBreak: A Scalable and Diverse Multi-turn Jailbreak Benchmark for Evaluating LLM Safety

Jialin Song, Xiaodong Liu, Weiwei Yang, Wuyang Chen, Mingqian Feng, Xuekai Zhu, Jianfeng Gao

ICML 2026.

Paper
SEMA: Simple yet Effective Learning for Multi-Turn Jailbreak Attacks

SEMA: Simple yet Effective Learning for Multi-Turn Jailbreak Attacks

Mingqian Feng, Xiaodong Liu, Weiwei Yang, Jialin Song, Xuekai Zhu, Chenliang Xu, Jianfeng Gao

ICLR 2026.

Paper Code
FLOPS: Forward Learning with OPtimal Sampling

FLOPS: Forward Learning with OPtimal Sampling

Tao Ren*, Zishi Zhang*, Jinyang Jiang, Guanghao Li, Zeliang Zhang, Mingqian Feng, Yijie Peng

ICLR 2025.

Paper Code
Do More Details Always Introduce More Hallucinations in LVLM-based Image Captioning?

Do More Details Always Introduce More Hallucinations in LVLM-based Image Captioning?

Mingqian Feng, Yunlong Tang, Zeliang Zhang, Chenliang Xu

Preprint arXiv 2024.

Paper
Will the Inclusion of Generated Data Amplify Bias Across Generations in Future Image Classification Models?

Will the Inclusion of Generated Data Amplify Bias Across Generations in Future Image Classification Models?

Zeliang Zhang, Xin Liang, Mingqian Feng, Susan Liang, Chenliang Xu

Preprint arXiv 2024.

Paper
Empowering LLMs with Pseudo-Untrimmed Videos for Audio-Visual Temporal Understanding

Empowering LLMs with Pseudo-Untrimmed Videos for Audio-Visual Temporal Understanding

Yunlong Tang, Daiki Shimada, Jing Bi, Mingqian Feng, Hang Hua, Chenliang Xu

AAAI 2025.

Paper
Video Understanding with Large Language Models: A Survey

Video Understanding with Large Language Models: A Survey

Yunlong Tang, Jing Bi, Siting Xu, …, Mingqian Feng, …, Jiebo Luo, Chenliang Xu

Preprint arXiv 2024.

Paper
Can CLIP Count Stars? An Empirical Study on Quantity Bias in CLIP

Can CLIP Count Stars? An Empirical Study on Quantity Bias in CLIP

Zeliang Zhang, Zhuo Liu, Mingqian Feng, Chenliang Xu

Findings of EMNLP 2024.

Paper
Forward Learning with Differential Privacy

Forward Learning with Differential Privacy

Mingqian Feng, Jinyang Jiang, Zeliang Zhang, Chenliang Xu

Forward Learning for Gradient-based Black-box Saliency Map Generation

Forward Learning for Gradient-based Black-box Saliency Map Generation

Mingqian Feng*, Zeliang Zhang*, Jinyang Jiang, Rongyi Zhu, Yijie Peng, Chenliang Xu

Preprint arXiv 2024.

Paper
Discover and Mitigate Multiple Biased Subgroups in Image Classifiers

Discover and Mitigate Multiple Biased Subgroups in Image Classifiers

Mingqian Feng*, Zeliang Zhang*, Jinyang Jiang, Rongyi Zhu, Yijie Peng, Chenliang Xu

CVPR 2024.

Paper Code

Education

Experiences

Misc.

Fun facts:

Pipe dreams:

Mingqian Feng
Template from Martin Saveski