姓 名：周弘毅
职 称：助理教授
研究方向：强化学习，大语言模型，人工智能理论
教授课程：人工智能导论
E - mail： zhouhongyi@mail.sufe.edu.cn
电话：
研究领域
强化学习，大语言模型，人工智能理论，深度学习，半参数模型等
教育经历
2021年09月 - 2026年07月 统计学 博士 清华大学
2017年09月 - 2021年07月数学与应用数学学士清华大学
工作经历
2026年08月 – 至今 助理教授上海财经大学统计与数据科学学院
研究成果
* Equal Contribution
Publications:
Hongyi Zhou*, Jin Zhu*, Kai Ye, Erhan Xu, Ying Yang, Chengchun Shi. (2026). Learn-to-Distance: Distance Learning for Detecting LLM-Generated Text,
International Conference on Learning Representations(ICLR).
Xinyi Qi*, Kai Ye*, Chengchun Shi, Ying Yang, Jin Zhu, Hongyi Zhou. (2026). A Difference-in-Difference Approach to Detecting AI-Generated Images. IEEE Conference on Computer Vision and Pattern Recognition (CVPR).
Hongyi Zhou, Wenqing Su, Qixian Zhong, Ying Yang. (2025+). Semiparametric Inference for Functional Survival Models, Statistica Sinica.
Hongyi Zhou, Josiah Hanna, Jin Zhu, Ying Yang, Chengchun Shi. (2025). Demystifying the Paradox of Importance Sampling with an Estimated History-Dependent Behavior Policy in Off-Policy Evaluation, International Conference on Machine Learning (ICML).
Hongyi Zhou*, Jin Zhu*, Pingfan Su, Kai Ye, Ying Yang, Shakeel Gavioli-Akilagun, Chengchun Shi. (2025); AdaDetectGPT: Adaptive Detection of LLM-Generated Text with Statistical Guarantees, Conference on Neural Information Processing Systems (NeurIPS).
Erhan Xu*, Kai Ye*, Hongyi Zhou*, Luhan Zhu, Francesco Quinzan, Chengchun Shi. (2025). Doubly Robust Alignment for Large Language Models, Conference on Neural Information Processing Systems (NeurIPS).
Jin Zhu*, Jingyi Li*, Hongyi Zhou, Yinan Lin, Zhenhua Lin, Chengchun Shi. (2025). Balancing Interference and Correlation in Spatial Experimental Designs: A Causal Graph Cut Approach, International Conference on Machine Learning (ICML)
Li Wang*, Hongyi Zhou*, Weidong Ma, Ying Yang. (2024). Expected Projection-Averaging of Conditional Distribution Function-based Measures for Independence Test and Feature Screening, Journal of Multivariate Analysis, 205, 105378.
Fei Ye, Hongyi Zhou, Ying Yang. (2023). Asymptotic Properties of Relative Error Estimation for Accelerated Failure Time Model with Divergent Number of Parameters, Statistics and Its Interface, 17(1), 107–125.
Manuscripts:
Hongyi Zhou*, Jin Zhu*, Ying Yang, Chengchun Shi.(2026) Detecting LLM-Generated Text with Performance Guarantees.(JASA major revision)
Kai Ye*, Hongyi Zhou*, Jin Zhu*, Francesco Quinzan, Chengchun Shi. (2025). Robust Reinforcement Learning from Human Feedback for Large Language Models Fine-Tuning. (AOAS major revision)
Hongyi Zhou*, Kai Ye*, Erhan Xu, Jin Zhu, Ying Yang, Shijin Gong, Chengchun Shi.(2026) Demystifying Group Relative Policy Optimization: Its Policy Gradient is a U-Statistic. https://arxiv.org/abs/2603.01162.
Shijin Gong*, Kai Ye*,Jin Zhu,Xinyu Zhang, Hongyi Zhou, Chengchun Shi. (2026). Kernelized Advantage Estimation: From Nonparametric Statistics to LLM Reasoning.https://arxiv.org/pdf/2604.28005.
奖励、荣誉
第二十六届京津冀青年概率统计学术研讨会“钟家庆优秀论文奖”2025
