I am a PhD student in Computer Science at the University of North Carolina at Chapel Hill. My research develops efficient, robust, and interpretable generative models, with a particular focus on recommendation and multimodal reasoning systems.
Curriculum Vitae Google Scholar
Education
- University of North Carolina at Chapel Hill — PhD in Computer Science, expected May 2029
- Columbia University — BS in Computer Science, August 2022–May 2024
Research Interests
Experience & News
- May 2026–Present: Research Scientist Intern, TikTok Feed Recommendation at ByteDance, developing long-sequence recommendation pipelines and KV-cache optimization.
- July 2026: Released Diffusion-GR2, a block-diffusion generative reasoning re-ranker developed with UNC and Meta.
- November 2025: Sparse Neurons Carry Strong Signals of Question Ambiguity in LLMs appeared at EMNLP 2025.
- July 2025: The Koala Benchmarks for the Shell received the USENIX ATC 2025 Best Paper Award.
- January 2025: GuideLLM was accepted to NAACL 2025.
Publications
Diffusion-GR2: Diffusion Generative Reasoning Re-ranker
Zhuoxuan Zhang*, Kangqi Ni*, Yuhang Chen*, Mingfu Liang*, Xiaohan Wei, Yunchen Pu, Fei Tian, Chonglin Sun, Frank Shyu, Adam Song, Sandeep Pandey, Luke Simon, Tianlong Chen, Xi Liu
arXiv 2026.
Sparse Neurons Carry Strong Signals of Question Ambiguity in LLMs
Zhuoxuan Zhang, Jinhao Duan, Edward Kim, Kaidi Xu
EMNLP 2025.
The Koala Benchmarks for the Shell: Characterization and Implications
Evangelos Lamprou, Ethan Williams, Georgios Kaoukis, Zhuoxuan Zhang, Michael Greenberg, Konstantinos Kallas, Lukas Lazarek, Nikos Vasilakis
USENIX ATC 2025, Best Paper Award.
GuideLLM: Exploring LLM-Guided Conversation with Applications in Autobiography Interviewing
Zhuoxuan Zhang*, Jinhao Duan*, Xinyu Zhao*, Eunhye Grace Ko, Lily Boddy, Chenan Wang, Tianhao Li, Alexander Rasgon, Junyuan Hong, Min Kyung Lee, Chenxi Yuan, Qi Long, Ying Ding, Tianlong Chen, Kaidi Xu
NAACL 2025.
An Exploration of LLM-Guided Conversation in Reminiscence Therapy
Zhuoxuan Zhang*, Jinhao Duan*, Xinyu Zhao*, Eunhye Grace Ko, Lily Boddy, Chenan Wang, Tianhao Li, Alexander Rasgon, Junyuan Hong, Min Kyung Lee, Chenxi Yuan, Qi Long, Ying Ding, Tianlong Chen, Kaidi Xu
GenAI4Health at NeurIPS 2024, Best Demo Paper Award.
Manuscripts Under Review
- Scaling Laws for Multimodal Large Reasoning Models with Arbitrary Modality Combinations
Jie Peng, Zhuoxuan Zhang, Gengwei Zhang, Huaizhi Qu, Hossein Nourkhiz Mahjoub, Vaishnav Tadiparthi, Kwonjoon Lee, Ehsan Moradi Pari, Yanyong Zhang, Tianlong Chen - Targeted Poisoning of Generative Recommendation Models via Context-Aligned Interactions
Zhuoxuan Zhang, Jinhao Duan, Xinyu Zhao, Jinduo Guo, Shuai Shao, Rui Li, Luke Simon, Zhun Deng, Tianlong Chen, Bo Sun - Faithful Chains: Enforcing Temporal Consistency in CoT Reasoning via Bellman Regularization
Zhuoxuan Zhang, Zhen Tan, Dawei Li, Bohan Jiang, Song Wang, Huan Liu, Tianlong Chen
* Equal contribution. Publication and manuscript information updated from the supplied CV and Google Scholar profile.
Miscellaneous
I love art in all its forms, both auditory and visual. Playing the violin is one of my greatest joys, and I have a deep appreciation for Western classical music. I also enjoy Peking opera—especially the Shang School Dan role—xiangsheng (traditional Chinese crosstalk), and classical Chinese dance, including Korean ethnic dance in the Chinese classical dance style.
I am also deeply interested in aviation. I enjoy learning about flight and am currently working toward my Private Pilot License (PPL).