🎉 Our paper "Towards Natural Machine Unlearning" is accepted by TPAMI
The final version will be released soon.
👨🏻💻About Me
Currently, I am a fourth-year Ph.D. student (Zhiyuan Honors) at Shanghai Jiao Tong University, studying under the supervision of Prof. Xiaolin Huang. My research interests primarily lie in efficient adaptation methods for LLMs or VLMs, with a specific focus on continual learning and their applications in intelligent agents. I welcome academic discussions and potential collaborations.
Interests
- Continual Learning
- Efficient Fine-tuning
- LLM-based Agent
Education
Ph.D. Student Automation Science and Engineering
Shanghai Jiao Tong University
BSc Automation
Shanghai Jiao Tong University
Featured Publications
Recent Publications
(2025).
Bi-LoRA: Efficient Sharpness-Aware Minimization for Fine-Tuning Large-Scale Models.
In arXiv.
(2025).
T2I-ConBench: Text-to-Image Benchmark for Continual Post-training.
In arXiv.
(2025).
A Unified Gradient-based Framework for Task-agnostic Continual Learning-Unlearning.
In arXiv.
(2025).
Simulating Training Dynamics to Reconstruct Training Data from Deep Neural Networks.
In ICLR 2025 Poster.
Recent News
🎉 Our paper "Simulating Training Dynamics to Reconstruct Training Data from Deep Neural Networks" is accepted by ICLR 2025
The final version and the code will be released soon.
🎉 Our paper "Unified Gradient-Based Machine Unlearning with Remain Geometry Enhancement" is accepted by NeurIPS 2024 as Spotlight
The final version and the code will be released soon.
Experience
Intern Algorithm Engineer
HuaweiResearch on Continual Post-training for Text-to-Image ModelsIntern Algorithm Engineer
ByteDanceAdvertising Business Algorithm Development
Education
Ph.D. Student Automation Science and Engineering
Shanghai Jiao Tong UniversityBSc Automation
Shanghai Jiao Tong University
Selected Projects
I enjoy making things. Here are a selection of projects that I have worked on over the years.
