Zhiyun Zhang

Hi! I’m a Master’s student in Machine Learning at Carnegie Mellon University. I’m currently interning as a Machine Learning Engineer at TikTok’s Recommendation Group.

My research interests lie in LLM reasoning, healthcare AI, and multimodal learning.

At CMU, I am fortunate to work in Prof. Chenyan Xiong’s lab, where my independent study focuses on training LLMs for faithful, evidence-based medical reasoning. Previously, I was a visiting student in Prof. Rose Yu’s lab at UC San Diego, working on faster and more reliable reverse latent space optimization for SO(3)-equivariant 3D small molecule generation.

Prior to CMU, I received my Bachelor’s degree with honors in Computer Science and Technology from Zhejiang University, where I was advised by Prof. Jie Yang at Harvard on large-scale clinical NLP datasets. I also interned at Alibaba DAMO Academy, working on multimodal pretraining and inference for medical data under limited modality settings.

news

May 18, 2026	Started my Machine Learning Engineer internship at TikTok Recommendation Group!
Feb 06, 2025	Started my Research Internship at Alibaba DAMO academy!

selected publications

EMNLP

FaithMed: Training LLMs for Faithful Evidence-Based Medical Reasoning

Zhiyun Zhang, Liwen Sun, Xiang Qian, and 1 more author

In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2026

Bib PDF Code

@inproceedings{zhang2026faithmed,
  title = {{FaithMed}: Training {LLMs} for Faithful Evidence-Based Medical Reasoning},
  author = {Zhang, Zhiyun and Sun, Liwen and Qian, Xiang and Xiong, Chenyan},
  booktitle = {Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing ({EMNLP})},
  year = {2026},
}

AAAI

Disentangling for Transfer: Boosting Limited Modalities via Information-Theoretic Regularization and Cross-Modal Reconstruction

Zhiyun Zhang, Yanjun Zhou, Yu Hu, and 5 more authors

In Proceedings of the AAAI Conference on Artificial Intelligence, 2026

Bib PDF

@inproceedings{zhang2026disentangling,
  title = {Disentangling for Transfer: Boosting Limited Modalities via Information-Theoretic Regularization and Cross-Modal Reconstruction},
  author = {Zhang, Zhiyun and Zhou, Yanjun and Hu, Yu and Ma, Xin and Yuan, Zheng and Wang, Zheng and Zhang, Hao and Xu, Min},
  booktitle = {Proceedings of the AAAI Conference on Artificial Intelligence},
  volume = {40},
  number = {15},
  year = {2026},
}

arXiv

GENERator: A Long-Context Generative Genomic Foundation Model

W. Wu, Q. Li, Y. Zhang, and 9 more authors

arXiv preprint arXiv:2502.07272, 2025

Bib PDF

@article{wu2025generator,
  title = {{GENERator}: A Long-Context Generative Genomic Foundation Model},
  author = {Wu, W. and Li, Q. and Zhang, Y. and Zhan, Z. and Chen, R. and Li, M. and Fu, K. and Qi, J. and Bao, Y. and Wang, C. and Zhang, Zhiyun and others},
  journal = {arXiv preprint arXiv:2502.07272},
  year = {2025},
}