Theoretical Physics | Computational Modeling | Machine Learning

Currently a first-year MSci Physics student at University College London (UCL). I bridge rigorous theoretical physics with advanced computational techniques. Rather than conforming to standard undergraduate pacing, I independently master graduate-level physics texts and architect machine learning models to tackle complex systems.

🎯 Actively seeking Summer Research Internships and Transfer Opportunities to top-tier physics programs.

• Focus: Graduate-Level Derivations, Analytical Mechanics, Quantum Mechanics
• Overview: The "hardcore" theoretical engine of my self-study, extending far beyond the standard Year 1 curriculum.
• Key Implementations:
  • A rigorous, dynamic repository of my independent mathematical derivations.
  • Deep-dive theoretical notes, formula derivations, and problem-solving based on absolute foundational texts: J.J. Sakurai's Modern Quantum Mechanics and Landau & Lifshitz's Mechanics.

• Tech Stack: Python, TensorFlow, Keras, Local LLMs (Ollama), NumPy
• Overview: My core engineering laboratory bridging algorithmic logic with physics data analysis.
• Key Implementations:
  • Architected and trained custom CNN & MLP architectures from scratch for high-accuracy pattern recognition.
  • Developed an autonomous Local LLM (Llama3) workflow agent for semantic physics archive management.
  • Programmed finite-difference solvers for fluid dynamics and conducted thermodynamic heat flux modeling.
  • Engineered a real-time, multimodal conversational AI assistant (Project Amadeus).

• Focus: Core Freshman Curriculum, Mathematical Methods, Laboratory Reports
• Overview: The solid academic baseline. A meticulously organized archive of my coursework for the MSci Physics program at University College London.
• Key Implementations:
  • Comprehensive documentation of standard university modules to maintain top-tier academic rigor and a flawless GPA foundation.
  • Structured notes covering introductory mechanics, electromagnetism, and practical experimental physics data.

• Tech Stack: Python, TensorFlow, Keras, NumPy
• Architected, trained, and optimized custom Convolutional Neural Networks (CNN) and Multi-Layer Perceptrons (MLP) from scratch for high-accuracy image classification.
• Focused on bridging algorithmic logic with data-driven analysis.

• Tech Stack: Python, Git Automation
• Engineered a custom local automation script that dynamically monitors file modifications in my physics datasets and automatically commits/pushes updates to GitHub, showcasing strong software engineering and workflow optimization capabilities.

• ☢️ CFD Modeling for HTGR: Spearheading Computational Fluid Dynamics simulations for next-generation High-Temperature Gas-cooled Reactor models within an independent water group (Rolls-Royce research program).
• 🪐 Thermodynamic Modeling of Europa: Formulated a mathematical model to quantify subsurface heat flux generated by orbital eccentricity and tidal dissipation, culminating in an independent research paper.(During high school)
• 📡 Quantum Communication Research: Authored and presented "Controlled remote implementation of operations for many systems" at CONF-MPCS 2024.(During high school)

• Languages: Python (Advanced), Java, LaTeX
• ML & Data: TensorFlow, Keras, NumPy, Matplotlib
• Engineering & Tools: Git CI/CD, Basic CFD Modeling, Overleaf
• Core Physics: Quantum Mechanics, Classical Mechanics, Statistical Thermodynamics

📫 Contact & Connect: * Email: liuchengxu2025@163.com

• ORCID: [Insert Your ORCID Link Here]