Hello, I’m Ganlin.
I am a rising junior in physics at Renmin University of China. I love using physics to understand the regularities behind the world around us, and I am equally fascinated by the mechanisms through which new capabilities and behaviors emerge in large language models.
I also enjoy conversations that move beyond equations: social science, philosophy, and the patterns and tensions we notice in everyday society. For me, these interests share the same starting point—a desire to observe carefully, build better explanations, and remain willing to revise them.
Research interests
2D Materials & Condensed Matter Physics — I am interested in the synthesis and characterization of two-dimensional materials. My current work explores the growth of transition-metal dichalcogenides, such as MoS₂, inside optical fibers using all-liquid-phase precursors.
AI for Science & LLM Emergence — I explore deep learning, generative AI, and multimodal language models for scientific work. I am particularly interested in parameter-efficient fine-tuning, over-parameterized adaptation, and how apparently new behaviors emerge from scale, data, and learning dynamics.
From material growth to characterization
Learning what control, reproducibility, and useful measurement mean inside a real experimental system.
From tuning models to understanding them
Asking why efficient adaptation works and how model capabilities genuinely connect with scientific tasks.
Connecting mathematical structures
Tracing shared ideas across special functions, eigenvalue problems, numerical optimization, and physical models.
Knowledge & tools
My physics background includes classical and analytical mechanics, electromagnetism, optics, and atomic physics, supported by linear algebra and mathematical methods for physics. I have worked with numerical techniques including Gaussian quadrature, Simpson’s rule, Monte Carlo integration, Newton methods, conjugate gradients, BFGS, and Lanczos algorithms.
I use Python, MATLAB, and PyTorch for computation and model experiments. Markdown, LaTeX, and Obsidian form the backbone of my personal knowledge system; this blog is the carefully edited public layer of that evolving archive.
Beyond physics
I read and talk about social science, philosophy, international relations, and contemporary social phenomena. I also value campus life and collaborative work, including organizing student activities. Experiments, code, and society are not separate rooms to me—they all train the habits of observing, modeling, collaborating, and correcting one’s judgment.
About this blog
This is my personal blog, updated every semester with study notes, derivations, physical intuition, and ideas that have become clearer through learning. The notes remain in Chinese because that is the language in which I study and annotate most naturally. They are living documents rather than finished textbooks.
The site is built with static HTML, CSS, and a small amount of JavaScript. Mathematical notation is rendered with KaTeX, and the site is deployed on Vercel. Note content is shared for learning and reference; please credit the source when quoting or reproducing it.