Zhuo Li
PhD CandidateEmail : lizhuo3 AT link.cuhk.edu.cn .Google Scholar
About
Hi, I am Zhuo Li , a 4-th year PhD student at The Chinese University of Hong Kong, Shenzhen , supervised by Dr. Xiang Wan and Prof. Tsung-Hui Chang . I obtained a B.E. degree in Data Science from the School of Computer Science and Engineering, Central South University, Changsha, China, in 2020.
Interests
My research interests lie in data-centric machine learning and its combinations with real-world applications. I am interested in data manipulation, probabilistic methods, and representation learning . These developed methods have been applied to robust long-tailed vision recognition, efficient data augmentation and weighting, sample selection and distillation, multimodal representation learning, and NLP. Recently, I am focused on developing methods for better alignment of LLM with downstream tasks and human preferences, such as automatic prompt engineering, data curation, reward modeling and RLHF algorithm.
News
08/2025 : Two paper accepted to EMNLP 2025.07/2025 : One paper accepted to TMLR 2025.07/2025 : One paper accepted to ACMMM 2025.05/2025 : One paper accepted to TMLR 2025.05/2025 : One paper accepted to ACL 2025 Oral .02/2025 : One paper accepted to TCSVT 2025.01/2025 : One paper accepted to NAACL 2025.
Publications
Human Preference Modeling, RLHF and Data Curation for LLM
APLOT: Robust Reward Modeling via Adaptive Preference Learning with Optimal Transport Zhuo Li , Yuege Feng, Dandan Guo, Jinpeng Hu, Anningzhe Gao, Xiang WanSimplify RLHF as Reward-Weighted SFT: A Variational Method Zhuo Li* , Pengyu Cheng*, Zhihong Chen, Yuejiao Xie, Xiang Wan, Anningzhe GaoAdd-One-In: Incremental Sample Selection for Large Language Models via a Choice-Based Greedy Paradigm Zhuo Li , Yuhao Du, Xiaoqi Jiao, Steven Y. Guo, Yuege Geng, Xiang Wan, Anningzhe Gao, Jinpeng Hu
Prompt Engineering and Task-Specific Modeling for NLP
Atoxia: Red-teaming Large Language Models with Target Toxic Answers Zhuo Li* , Pengyu Cheng, Xiang Wan, Anningzhe GaoSelf-Instructed Derived Prompt Generation Meets In-Context Learning: Unlocking New Potential of Black-Box LLMs Zhuo Li* , Yuhao Du*, Jinpeng Hu, Xiang Wan, Anningzhe GaoOral .A Simple yet Effective Subsequence-Enhanced Approach for Cross-Domain NER Zhuo Li , Zhihong Chen, Xiang Wan, Tsung-Hui ChangGraph Enhanced Contrastive Learning for Radiology Findings Summarization Zhuo Li* , Zhihong Chen, Zhen Li, Xiang Wan, Tsung-Hui Chang
Data Scarcity Scenarios
Synthesizing Minority Samples for Long-tailed Classification via Distribution Matching Zhuo Li , He Zhao, Zhen Li, Dandan Guo, Xiang Wan, Hongyuan ZhaPrototype-oriented Clean Subset Extraction for Noisy Long-tailed Classification Zhuo Li , He Zhao, Anningzhe Gao, Dandan Guo, Tsunghui Chang, Xiang WanBoosting Visual Feature Discrimination through Textual Prototypes in Imbalanced Data Scenarios Zhuo Li , He Zhao, Anningzhe Gao, Dandan Guo, Tsunghui Chang, Xiang WanEnhancing Minority Classes by Mixing: An Adaptive Optimal Transport Approach for Long-tailed Classification Zhuo Li , Dandan GuoLearning to re-weight examples with optimal transport for imbalanced classification Zhuo Li , Meixi Zheng, He Zhao, Mingyuan Zhou, Hongyuan ZhaDiverse Condensed Data Generation via Class Preserving Distribution Matching Zhuo Li , He Zhao, Mingyuan Zhou, Hongyuan Zha
Other Papers
Beyond Emotion Recognition: A Multi-Turn Multimodal Emotion Understanding and Reasoning Benchmark Zhuo Li , Peipei Song, Meng WangToward Explainable and Fine-Grained 3D Grounding through Referring Textual Phrases Zhuo Li* , Xuhao Li, Yao Guo, Shuguang Cui, Zhen LiCognitive-Visual Fusion for Target Classification in Rapid SAR Image Series Visual Presentation Zhuo Li* , Xiang Wan, Ruimao Zhang, Liangcheng Qu, Zhen Li, Kuiying Yin
Intern
Intern, Noah's Ark Lab, Huawei, May. 2025 - Now.
Intern, LIGHT STUDIO, Tencent, Dec. 2024 - Apr. 2025.
Teaching
Teaching Assistant, MFE5330 Regulatory Technology, Spring 2022-2025.
Teaching Assistant, MFE5231 Special Topic: Corporate Finance and Risk Management, Spring 2024-2025.
Teaching Assistant, MFE5190 Credit Risk Modeling and Products, Summer 2023-2025.
Service
Reviewer: NeurIPS, ICML, ICLR, IJCAI, ECCV, AISTATS, CVPR, ICCV, ACMMM, TNNLS.
Awards
Gold Class of Shenzhen Research Institute of Big Data PhD Fellowship 2021-2025.
3rd Place Award in semantic segmentation track of Urban3D@ECCV2022.
Last update: 03/2025. Template credited to Tengyu Ma 's and Danqi Chen 's.
