About me

Hi, I am Chengxun Jin.

I am currently an undergraduate student majoring in Artificial Intelligence at the College of Artificial Intelligence, Jilin University.

My current research interests include:

• Large language model fine-tuning
• Diffusion model fine-tuning
• Model quantization

This website is used to share my academic background, projects, publications, notes, and CV.

News

• 2026.05.09: I was invited by Tsinghua University to attend the 2026 Student Academic Exchange Day of the Institute for Artificial Intelligence, Tsinghua University, on May 15. See you at Tsinghua!
• 2026.05.01: Our paper, LOZO+: Provably Efficient Zeroth-Order Fine-Tuning via Greedy Low-Rank Subspace Selection, was accepted to ICML 2026 as a regular paper.

Research Interests

I hope to focus my future research on optimizing attention mechanisms, PEFT-based fine-tuning, and multimodal large models.

Contact

You can find me on GitHub: Megumin-forever