Biography
I am a second-year M.S. student at Institute of Media, Information, and Network (min), Shanghai Jiao Tong University, advised by Prof. Hongkai Xiong and Prof. Wenrui Dai. I received my bachelor’s degree in Electronic Science and Technology from SJTU.
I have a keen interest in machine learning and interpretable artificial intelligence. My goal is to identify the fundamental optimization goal of deep learning and construct interpretable machine learning systems through mathematical derivations. Specifically, I aim to establish objective metrics to measure model performance and mathematically deduce the factors that contribute to model performance, guiding the design of neural network architectures.
Download my CV.
- Machine Learning
- Interpretable AI
- Computer Vision
M.S. in Information and Communication Engineering, 2023
SJTU
B.S. in Electronic Science and Technology (major) and Computer Science and Technology (minor), 2019
SJTU
Publications
Research Experience
Advised by Prof. Hongkai Xiong and Prof. Wenrui Dai, my researches include:
- Proposed ChebyNet, a novel network paradigm to build Chebyshev polynomial connections between general network layers. Specifically, established recursive relatetionship among adjacent layers and polynomial relationship between non-adjacent layers. Comprehensive experiments verify its strong approximation capacity. (ICLR 2025 on submission)
- Proposed GLEAM, an efficient fine-tuning method for large model parameters. This method leverages the high similarity of parameter matrices in LoRa to construct a low-rank decomposition, further reducing the number of parameters required for fine-tuning while enhancing performance. (AAAI 2025 on submission)
Advised by Prof. Jieping Ye, my researches include:
- Proposed a new paradigm that calculates the importance of different components of few-shot CoT demonstrations using Shapley values, thereby enhancing the inference capability of large models. This approach elucidates the mechanism of CoT in large models and unifies previous research. (NAACL 2025 on submission)
Advised by Prof. Quanshi Zhang, my researches include:
- Theoretically derived the analytical solution for multi-step adversarial attacks, which explains the reasons behind the optimization difficulties in adversarial training. This is validated through experimental results. (Accepted by AAAI 2024)
- Theoretically derived the two-stage dynamic interaction process of DNNs, proving that the network learning process gradually encodes interactions of varying complexity. This provides a theoretical foundation for understanding overfitting. (Accepted by NeurIPS 2024)
- Theoretically derived and validated the robustness of concepts with different complexities.
Advised by Prof. Chongyang Zhang, my researches include:
- Developped Swin Transformer based model to implemente instance segmentation of workpiece welding area.
- Designed a space-time filter to remove false positive samples in pedestrian detection.
- Developped YOLOv5-based model to detect tower crane, recognize dangerous tower crane, and label electronic fence.
Competitions
Honors and Awards
Skills
Python, C++, Matlab, LaTeX, Linux, PyTorch, NumPy, Anaconda, Git, OpenCV
calculus, linear algebra, probability statistics
mandarin (native), English (fluent)
Activities and Leaderships
Contributions include:
- Third Place in the Team Category at the Tizong Cup in 2021.
- Second Place in the Team Category at the School Sports Meet in 2022.
Contributions include:
- Second Place in the Men’s 4$\times$100-Meter Relay at the School Sports Meet in 2020.
- First Place in the Men’s 4$\times$100-Meter Relay at the 2021 Track and Field Athletics Meet.