Zhuan Li

PhD candidate in Condensed Matter Physics

University of Pittsburgh, Pittsburgh

Biography

Zhuan is a Ph.D. candidate in Physics at the University of Pittsburgh, mentored by Prof. Roger Mong. He specializes in scientific computing, data analysis, and machine learning, applying advanced statistical methods and computational techniques to complex theoretical and practical challenges.

Currently as a summer graduate intern at Los Alamos National Lab under the supervision of Dr. Andrey Lokhov, Zhuan focuses on enhancing modeling systems and optimizing large-scale sampling operations. At the University of Pittsburgh, he focuses on condensed matter theory, especially the topological phase of matter and its potential application to quantum computing. Utilizing various computational libraries, Zhuan performs in-depth simulations and data analyses to tackle intricate problems across diverse fields, including quantum error correction, quantum information, and superconducting devices.

Before joining UPitt, Zhuan studied physics at the University of Chinese Academy of Sciences, China (BSc). During his undergraduate study, he worked on low-rank approximation algorithms based on the tensor network under the supervisor of Prof. Pan Zhang, Institute of Theoretical Physics, Chinese Academy of Sciences.

With a robust background in theoretical physics and applied statistics, Zhuan is well-prepared to contribute to data science and machine learning projects, aiming to drive technological and industrial advancements through innovative approaches.

Download my CV.

Interests

Statistical learning
Scientific Computing
Tensor network algorithm
Machine Learning
Quantum Computing

Education

PhD in Condensed Matter Physics, 2024
University of Pittsburgh, Pittsburgh, PA
BSc in Physics, 2019
University of Chinese Academy of Sciences, China
Visiting student, 2018
University of Bristol, UK

Projects

Credit Card Fraud Detection Using Machine Learning

In this project, I tackled the challenge of credit card fraud detection, an issue that affects financial institutions and consumers worldwide. The objective was to develop a robust machine learning system capable of identifying fraudulent transactions effectively. To to that, I employed three distinct models: logistic regression, random forest, and shallow neural networks. Each model was chosen for its unique strengths in handling binary classification problems and its potential to perform well on imbalanced datasets.

Galaxy Classification Using Machine Learning

This project aims to classify galaxies based on their morphologies using advanced machine learning techniques. In this project, ResNet50 is adapted for feature extraction, and K-means and spectral clustering were utilized to categorize galaxy images into four types: Cigar-shaped, Spiral, and two groups of Elliptical galaxies, one of which exhibited more noise. The analysis demonstrated the effectiveness of these techniques in distinguishing subtle differences in galaxy structures.

Meme Sentiment Analysis Using PyTorch

In this project, I developed a Multimodal Sentiment Analysis System using PyTorch that leverages both text and image data to analyze and predict the sentiment of memes. The primary objective was to construct a robust model capable of interpreting complex data inputs, including textual content from the memes and their corresponding images, to accurately determine the underlying sentiment, ranging across multiple categorical levels.

Researches

$\phi_0$-Josephson junction

A Josephson junction is a superconducting electronic device consisting of two superconducting electrodes separated by a thin insulating barrier. Current can flow through this junction without any resistance and is called the supercurrent.

$$\phi_0$-Josephson junction$

Tensor network and the entanglment of purification

In quantum mechanics, tensor network representations like the matrix product state (MPS) simplify the handling of complex wave functions by breaking them down into a series of linked tensors, each representing a system site. These tensors use auxiliary indices to represent entanglement, with more entangled systems requiring larger indices. For mixed states, the entanglement of purification (EP) indicates the necessary size of auxiliary indices for accurate representation, but is challenging to calculate precisely, so estimates of its bounds are used instead.

Topological order on a torus

The Fractional Quantum Hall effect revealed new phases of matter beyond traditional paradigms, characterized by ground state degeneracy, nontrivial particle statistics, and gapless edge excitations, making them promising for fault-tolerant quantum computation. Determining the topological phase from a microscopic Hamiltonian is challenging, as it often assumes the ground state contains all topological information. It is proposed that topological data can be extract from ground-state wavefunction overlaps, however, this method was impractical for lattice systems. This issue was addressed by constructing specific operators and managing phase factors, applying Wen’s results to lattice systems.

Publications

Zhuan Li, Roger Mong (2024). Detecting topological order from modular transformations of ground states on the torus.

PDF Cite

Nicholas M. Hougland, Zhuan Li, David Pekker (2024). Evidence of φ0-Josephson junction from skewed diffraction patterns in Sn-InSb nanowires.

PDF Cite

Bomin Zhang, Zhuan Li, Sergey Frolov (2023). Evidence of φ0-Josephson junction from skewed diffraction patterns in Sn-InSb nanowires.

PDF Cite

Zhuan Li, Roger Mong (2022). Detecting topological order from modular transformations of ground states on the torus. Phys. Rev. B.

PDF Cite DOI

Bomin Zhang, Zhuan Li, Sergey Frolov (2022). Evidence of φ0-Josephson junction from skewed diffraction patterns in Sn-InSb nanowires.

PDF Cite

Activities

Teaching

I have been a Teaching Assistant at University of Pittsburgh for the following course:

PHYS 0212 Introduction to laboratory physics, Fall 2019
PHYS 0212 Introduction to laboratory physics, Spring 2020,
PHYS 0111 Introduction to physics 2, Summer 2020
PHYS 0212 Introduction to laboratory physics, Spring 2021,
PHYS 0212 Introduction to laboratory physics, Spring 2022,

Conferences, Schools and Events

I have given contributed talks on the following conferences:

2022 APS March Meeting (Chicago)
- Contributed talk. Title: Detecting topological order from modular transformations of ground states on the torus.
2023 APS March Meeting (Las Vegas)
- Contributed talk. Title: Estimating the reflected entropy from random matrices.
2024 APS March Meeting (Minneapolis)
- Contributed talk. Title: Replica Topological Order and Error Correction.

Awards

2022 Thomas-Lain essay competition.
2023 Pittsburgh Quantum Institute Fellowship.

Experience

Summer student

Los Alamos National Lab

May 2024 – Present Los Alamos, NM, USA

Explore new methods for a high-fidelity sampling using quantum annealers under the supervision of Dr. Andrew Lokhov.

PhD student

University of Pittsburgh

Sep 2019 – Present Pittsburgh, PA, USA

PhD in Physics. Study topological order and tensor network under the supervision of Prof. Roger Mong.

Visiting student

University of Bristol

Jan 2018 – Jun 2018 Bristol, UK

Visiting student

Institute of Theoretical Physics, Chinese Academy of Sciences

Jun 2017 – Jan 2018 Beijing, China

Study low-rank approximation algorithms on tensor networks under the supervision of Prof. Pan Zhang.

BSc in Physics

University of Chinese Academy of Sciences

Sep 2015 – Jul 2019 Beijing, China