About Me

• Hi! I am a third-year Ph.D. student at EECS, University of Michigan, advised by Prof. Qing Qu. Before joining U-M, I received my undergraduate and master’s degrees from Wuhan University, advised by Prof. Weiwei Liu.
• Generative AI and Machine Learning. I work on the foundations and applications of generative models, including diffusion models, diffusion language models, generalization and memorization, and in-context learning in world models.
• Machine Learning. I also investigated out-of-domain generalization and adversarial robustness in AI systems, with the goal of building models that remain reliable under distribution shifts and adversarial attacks.

Recent News

• 5/14/2026: ForcingDAS (data assimilation via diffusion forcing) released on arXiv.
• 4/30/2026: Evaluating diffusion representations via self-supervision accepted by ICML 2026.
• 1/25/2026: Balanced representations for diffusion generalization accepted by ICLR 2026.
• 9/26/2025: Passed the qualifying exam and became a Ph.D. candidate.
• 9/18/2025: Model collapse through generalization vs. memorization accepted by NeurIPS 2025 (Spotlight).

Publications

A Closer Look at Model Collapse: From a Generalization-to-Memorization Perspective
Lianghe Shi*, Meng Wu*, Huijie Zhang, Zekai Zhang, Molei Tao, Qing Qu.
ICML 2025 Workshop (Oral) and NeurIPS 2025 (Spotlight).
Website

Evaluating the Representation Space of Diffusion Models via Self-Supervised Principles
Xiao Li*, Yixuan Jia*, Zekai Zhang, Xiang Li, Lianghe Shi, Jinxin Zhou, Zhihui Zhu, Liyue Shen, Qing Qu.
ICML 2026.

Generalization of Diffusion Models Arises with a Balanced Representation Space
Zekai Zhang*, Xiao Li*, Xiang Li, Lianghe Shi, Meng Wu, Molei Tao, Qing Qu.
ICLR 2026 and DeepMath (Oral).

ForcingDAS: Unified and Robust Data Assimilation via Diffusion Forcing
Yixuan Jia, Siyi Chen, Yida Pan, Xiao Li, Lianghe Shi, Chanyong Jung, Haijie Yuan, Ismail Alkhouri, Yue Cynthia Wu, Saiprasad Ravishankar, Jeffrey A. Fessler, Qing Qu.
arXiv preprint.

Adversarially robust unsupervised domain adaptation
Lianghe Shi*, Weiwei Liu. The journal of Artificial Intelligence. AIJ.

A Closer Look at Curriculum Adversarial Training: From an Online Perspective
Lianghe Shi*, Weiwei Liu. Proceedings of the 38th AAAI Conference on Artificial Intelligence. AAAI 2024.

Adversarial self-training improves robustness and generalization for gradual domain adaptation
Lianghe Shi*, Weiwei Liu. 37th Conference on Neural Information Processing Systems. NeurIPS 2023.

Teaching

• Graduate Student Instructor of EECS 453: Principles of Machine Learning (Fall 2026)
• Graduate Student Instructor of EECS 501: Probability and Random Processes (Fall 2025)

Academic Services

• Conference Reviewer: NeurIPS, ICML, ICLR, AISTATS.