Publications

Preprints

* represents equal contribution

• Yihua Zhang*, Hongkang Li*, Yuguang Yao*, Aochuan Chen, Shuai Zhang, Pin-Yu Chen, Meng Wang, Sijia Liu. “Visual Prompting Reimagined: The Power of Activation Prompts.”

Conference Papers

* represents equal contribution

• Hongkang Li, Meng Wang, Songtao Lu, Xiaodong Cui, Pin-Yu Chen. Training Nonlinear Transformers for Chain-of-Thought Inference: A Theoretical Generalization Analysis, ICLR 2025.
• Hongkang Li, Yihua Zhang, Shuai Zhang, Meng Wang, Sijia Liu, Pin-Yu Chen. When is Task Vector Provably Effective for Model Editing? A Generalization Analysis of Nonlinear Transformers, ICLR 2025 🔥 Oral.
• Yuankai Luo, Hongkang Li, Qijiong Liu, Lei Shi, Xiao-Ming Wu. Node Identifiers: Compact, Discrete Representations for Efficient Graph Learning, ICLR 2025.
• Yuankai Luo, Hongkang Li, Lei Shi, Xiao-Ming Wu. Enhancing Graph Transformers with Hierarchical Distance Structural Encoding, Neurips 2024.
• Hongkang Li, Meng Wang, Songtao Lu, Xiaodong Cui, Pin-Yu Chen. How Do Nonlinear Transformers Learn and Generalize in In-Context Learning?, ICML 2024.
• Hongkang Li, Meng Wang, Tengfei Ma, Sijia Liu, Zaixi Zhang, Pin-Yu Chen. What Improves the Generalization of Graph Transformers? A Theoretical Dive into the Self-attention and Positional Encoding, ICML 2024.

• Hui Wan, Hongkang Li, Songtao Lu, Xiaodong Cui, Marina Danilevsky. How Can Personalized Context Help? Exploring Joint Retrieval of Passage and Personalized Context, ICASSP 2024.

• Shuai Zhang, Hongkang Li, Meng Wang, Miao Liu, Pin-Yu Chen, Songtao Lu, Sijia Liu, Keerthiram Murugesan, Subhajit Chaudhury. On the Convergence and Sample Complexity Analysis of Deep Q-Networks with -Greedy Exploration, Neurips 2023.

• Hongkang Li, Meng Wang, Sijia Liu, Pin-Yu Chen. A Theoretical Understanding of shallow Vision Transformers: Learning, Generalization, and Sample Complexity, ICLR 2023. Poster

• Hongkang Li, Meng Wang, Sijia Liu, Pin-Yu Chen, Jinjun Xiong. Generalization Guarantee of Training Graph Convolutional Networks with Graph Topology Sampling, ICML 2022. Poster
• Hongkang Li, Shuai Zhang, Meng Wang. Learning and generalization of one-hidden-layer neural networks, going beyond standard gaussian data, CISS 2022. Poster

Journal Papers

* represents equal contribution

• Jiawei Sun, Hongkang Li, Meng Wang. Theoretical Learning Performance of Graph Networks: the Impact of Jumping Connections and Layer-wise Sparsification, Transaction on Machine Learning Research, May 2025.

• Hongkang Li, Shuai Zhang, Yihua Zhang, Meng Wang, Sijia Liu, Pin-Yu Chen. How does promoting the minority fraction affect generalization? A theoretical study of the one-hidden-layer neural network on group imbalance, IEEE Journal of Selected Topics in Signal Processing (JSTSP), Special Series on AI in Signal & Data Science - Toward Explainable, Reliable, and Sustainable Machine Learning, March 2024.

Workshop Papers

* represents equal contribution

• Jiawei Sun, Shuai Zhang, Hongkang Li, Meng Wang. Theoretical Guarantees and Training Dynamics of Contrastive Learning: How Misaligned Data Influence Feature Purity, (HiLD 2024) 2nd Workshop on High-dimensional Learning Dynamics at ICML 2024.

• Hongkang Li, Songtao Lu, Xiaodong Cui, Pin-Yu Chen, Meng Wang. Understanding Mamba in In-Context Learning with Outliers: A Theoretical Generalization Analysis, (HiLD 2024) 2nd Workshop on High-dimensional Learning Dynamics at ICML 2024.

• Hongkang Li, Meng Wang, Shuai Zhang, Sijia Liu, Pin-Yu Chen. Learning on Transformers is Provable Low-Rank and Sparse: A One-layer Analysis, (SAM 2024) 2024 IEEE 13th Sensor Array and Multichannel Signal Processing Workshop.

• Hongkang Li, Meng Wang, Songtao Lu, Hui Wan, Xiaodong Cui, Pin-Yu Chen. Transformers as Multi-Task Feature Selectors: Generalization Analysis of In-Context Learning, (M3L 2023) Mathematics of Modern Machine Learning Workshop at NeurIPS 2023.

Thesis

Hongkang Li. Optimization and Generalization Analysis of Advanced Neural Networks and Learning Algorithms, 2024.