CV

Curriculum Vitae

Personal Information

Visa Status: F-1
Private email: chan.li.physics@gmail.com
Phone: +1-858-370-9516

Present Position

Ph.D. Student in Theoretical Physics, University of California San Diego (2023–present)
Scheduled to advance to Ph.D. candidacy in December 2025.

Work Details

Address: Room 7258, Urey Hall, University of California San Diego,
Department of Physics, 9500 Gilman Drive, La Jolla, CA 92093, USA
Professional email: chanli@ucsd.edu

Education

University of California San Diego (UCSD) — La Jolla, USA
Ph.D. in Physics (anticipated June 2028)
Advancement to candidacy scheduled December 2025; GPA: 3.96/4.0
September 2023 – Present

Sun Yat-sen University — Guangzhou, China
M.Sc. in Theoretical Physics, GPA: 89/100 (3.55/4.0)
August 2020 – June 2023

Sun Yat-sen University — Guangzhou, China
B.Sc. in Optoelectronic Information Science and Engineering, GPA: 88/100 (3.78/4.0)
August 2016 – June 2020

Selected Honors and Awards

Physics Excellence Award, UC San Diego (Top 10%) — Sep 2023
Chinese National Scholarship, Ministry of Education, P.R. China (Top 1%) — Oct 2019; Oct 2021
Chen-Ning Franklin Yang Scholarship (Graduate Physics), Sun Yat-sen University (Top 1%) — Nov 2021
University First-Class Scholarship for Outstanding Students, Sun Yat-sen University (Top 10%) — Oct 2019
3rd Prize, Three-Minute Video Presentation Contest, IOP Publishing & X-MOL (Top 10%) — May 2021

Research Interests

Statistical and probabilistic approaches to understanding learning dynamics and generalization in neural networks.
Bayesian and statistical modeling for interpretable and reliable machine learning systems.
Non-equilibrium and critical phenomena in the optimization and training dynamics of overparameterized models.
Developing scalable and efficient algorithms that integrate theoretical insight with modern deep learning and AI-driven analytics.
Advancing transparent, data-efficient, and theory-guided methodologies for intelligent decision-making and predictive modeling.

Work Experience

Graduate Student Researcher — University of California San Diego
Spring 2024 – Present
Supervisor: Prof. Nigel Goldenfeld

Conduct research on statistical physics and machine learning, focusing on learning dynamics, generalization, and scaling behavior in neural networks.
Develop and maintain computational models and simulation pipelines to analyze large-scale learning systems.

Teaching Assistant — University of California San Diego
Winter 2024
Supervisor: Dr. Jessica Arlett

Assisted in teaching undergraduate physics courses, including leading discussion sections and grading assignments.
Supported students through office hours and contributed to developing problem sets and instructional materials.

Research Experience

University of California San Diego (UCSD) — La Jolla, USA

Ph.D. Researcher in Physics | Advisor: Prof. Nigel Goldenfeld

Project 1: Learning Dynamics and Scaling Behaviors in Large Neural Networks (Oct 2024 – Oct 2025)
Developed physics-inspired models to explain robust generalization in overparameterized neural networks, informing scalable and reliable AI design.
Under review at Physical Review Letters (top-tier peer-reviewed journal).
Project 2: Learning Algorithms for Emergent Symmetry Discovery in Physical Systems (Apr 2025 – Present)
Designed learning algorithms to identify hidden symmetries and invariant structures from physical data.
Combined theoretical analysis with data-driven modeling to enhance interpretability. (Manuscript in preparation.)

Sun Yat-sen University — Guangzhou, China

Researcher in PMI Lab | Advisor: Prof. Haiping Huang

Project 1: Interpretable Learning Framework for Neural Networks (Aug 2019 – Oct 2021)
Developed probabilistic and mean-field learning models explaining how deep and recurrent neural networks coordinate parameters.
Published in Phys. Rev. Lett. 125, 178301 (2020) and Phys. Rev. E 107, 024307 (2023).
Project 2: Hierarchical Representations in Deep Learning (Aug 2021 – Oct 2022)
Designed algorithms decomposing neural network weights into interpretable latent modes.
Published in Phys. Rev. Research 5, L022011 (2023).
Project 3: Continual and Multi-Task Learning Frameworks (Aug 2020 – Oct 2022)
Developed physics-inspired frameworks mitigating catastrophic forgetting and enhancing knowledge transfer.
Published in Phys. Rev. E 108, 014309 (2023).
Project 4: Predictive Coding Models for Language Processing (Apr 2023 – Apr 2024)
Implemented predictive coding models to study information flow and adaptive learning in neural circuits.
Published in Phys. Rev. E 109, 044309 (2024).

Professional Skills

Programming and Software:
Proficient in Python for machine learning, scientific computing, and data analysis;
experienced with PyTorch, NumPy, SciPy, and Matplotlib;
familiar with C/C++, Linux, and LaTeX; working knowledge of Matlab and Mathematica.

Machine Learning and Optimization:
Designing and training neural networks, implementing optimization algorithms (SGD, Adam), and analyzing training dynamics, generalization, and scaling behaviors.

Statistical and Probabilistic Modeling:
Strong foundation in statistical physics, Bayesian inference, and probabilistic modeling, including mean-field and random matrix analysis.

Communication:
Effective in cross-disciplinary collaboration and clear presentation of complex technical ideas across physics, computer science, and applied mathematics.

Languages:
Native in Chinese; fluent in English (IELTS 7.5).

Publications (peer-reviewed, comparable to top-tier AI/ML conferences)

Learning credit assignment

Chan Li,Haiping Huang. (2020). "Learning credit assignment." Physical Review Letters.

Ensemble perspective for understanding temporal credit assignment

Wenxuan Zou1,*, Chan Li1,*, and Haiping Huang. (2023). "Ensemble perspective for understanding temporal credit assignment." Physical Review E.

Emergence of hierarchical modes from deep learning

Chan Li,Haiping Huang. (2023). "Emergence of hierarchical modes from deep learning." Physical Review Research.

Statistical mechanics of continual learning: Variational principle and mean-field potential

Chan Li1,*, Zhenye Huang2,*, Wenxuan Zou1,*, and Haiping Huang. (2023). "Statistical mechanics of continual learning: Variational principle and mean-field potential." Physical Review E.

Meta predictive learning model of languages in neural circuits

Chan Li1 and Haiping Huang. (2024). "Meta predictive learning model of languages in neural circuits." Physical Review E.

Chan Li

CV