About me
Welcome! I am currently a first-year Ph.D. student of computer science at University of Pittsburgh, advised by Prof. Junyu Liu. Before that, I received my B.E. from Tianjin University, and my M.S. degree from Georgia Institute of Technology.
I work on large language models (dynamics, and the applications in quantum), quantum learning theory and applying physical principles to analyze neural network dynamics. I am interested in leveraging principles or tools from physics for AI dynamics (e.g., fine-tuning dynamics, grokking, emergent abilities, neural scaling law, catapult dynamics, etc.)/applications and AI for quantum.
I am actively seeking LLM or Quantum internship positions in the US for Summerβ26.
Feel free to reach out for questions and collaborations!
Selected Publications
Artificial Entanglement in the Fine-Tuning of Large Language Models
Min Chen, Zihan Wang, Canyu Chen, Zeguan Wu, Manling Li, Junyu Liu
arXiv preprint
π Arxiv
Information Bottleneck-guided MLPs for Robust Spatial-Temporal Forecasting
Min Chen, Guansong Pang, Wenjun Wang, Cheng Yan
ICML 2025
π Openreview | π» Code
Catapult Dynamics and Phase Transitions in Quadratic Nets
David Meltzer, Min Chen, Junyu Liu
J. Stat. Mech.: Theory Exp. (JSTAT)
π Arxiv | π IOPScience
GroverGPT-2: Simulating Groverβs Algorithm via Chain-of-Thought Reasoning and Quantum-Native Tokenization
Min Chen, Jinglei Cheng, Pingzhi Li, Haoran Wang, Tianlong Chen, Junyu Liu
Submit to npj Quantum Information (under peer review)
π Arxiv | π» Code
An Analytic Theory of Quantum Imaginary Time Evolution
Min Chen, Bingzhi Zhang, Quntao Zhuang, Junyu Liu
arXiv preprint
π Arxiv
Awards & Grants
- Tinker Research Grant, 2025, $5,000 USD, Thinking Machines Lab
Competitive research grant supporting independent research with compute credits and priority technical support.
Internships
- 2024.08 - 2024.12, Research Assistant, University of Macau, Macau, China
Direction: System for LLM Serving
Supervised by Dr. Huanle Xu
Services
- Reviewer: KDD 2024, SLLM@ICLR 2025, npj Quantum Information
