CV简历
Curriculum vitae of Moxian Qian.钱莫闲的简历。
Contact Information联系方式
| Name姓名 | Moxian Qian |
| Title身份 | MSc Student硕士研究生 |
| Email邮箱 | mqian@students.uni-mainz.de |
| Location地址 | Staudingerweg 18, Mainz, Rhineland-Palatinate 55128 |
| Profiles学术主页 | ORCID (0009-0003-5919-5295) · InspireHEP (Qian Moxian) · ResearchGate (Moxian Qian) |
Professional Summary个人简介
MSc student in theoretical physics at Johannes Gutenberg University Mainz, currently based at the Helmholtz Institute Mainz, and affiliated with IQCB and IMB. My research includes machine-learning methods for lattice field theory and lattice QCD. Previously a research assistant at CUHK-Shenzhen (2022–2025).美因茨大学（JGU）理论物理硕士在读，目前在美因茨亥姆霍兹研究所（HIM），同时在 IQCB 与 IMB。研究方向包括格点场论 / 格点 QCD 的机器学习方法。此前在香港中文大学（深圳）做了大约三年（2022–2025）的 research assistant。
Experience工作经历
-
2022 - 2025 Shenzhen, China中国 深圳
Research AssistantResearch Assistant（科研实习）
The Chinese University of Hong Kong, Shenzhen (CUHK-Shenzhen)香港中文大学（深圳）
Roughly three years of research-assistant work between undergrad and graduate school, supervised by Prof. Jianhui Zhang and Prof. Kai Zhou. Where I started learning lattice field theory and machine-learning methods for physics.本科与硕士之间约三年的 research assistant 工作，由 张建辉 老师与 周凯 老师共同指导。在这里入门了格点场论以及「用机器学习做物理」。
- Advisors: Prof. Jianhui Zhang, Prof. Kai Zhou.导师：张建辉 老师、周凯 老师。
- Hands-on training in lattice field theory, scientific computing, and machine learning for physics.格点场论、科学计算、面向物理的机器学习等方向的实战训练。
- Computational algebraic geometry, plus integration-by-parts (IBP) and integrand-reduction methods for Feynman diagrams, applied to jet algorithms.计算代数几何，以及 Feynman 图的 IBP 与 integrand-reduction 方法，应用于 jet algorithm。
Education教育经历
-
2026 - present至今 Mainz, Germany德国 美因茨
M.Sc. (in progress)硕士在读
Johannes Gutenberg University Mainz美因茨大学（JGU）
Theoretical Physics — Lattice Field Theory & Machine Learning理论物理 — 格点场论 与 机器学习
- Based at the Helmholtz Institute Mainz. Also affiliated with IQCB and IMB.目前在美因茨亥姆霍兹研究所（HIM），同时在 IQCB 与 IMB。
- Working on machine-learning samplers for lattice field theory.做格点场论的机器学习采样器。
- Operator spectroscopy of trained normalizing-flow / autoregressive samplers.对训练后的 normalizing flow / 自回归采样器做算符谱学分析。
-
2017 - 2022 Hohhot, China中国 呼和浩特
B.Sc.理学学士
Inner Mongolia Normal University (IMNU)内蒙古师范大学
Physics (switched from Material Physics mid-program)物理学（中途从材料物理转入）
- Advisors: Prof. Xibin Li, Prof. Yongjun Cao.导师：李喜彬 老师、曹永军 老师。
- Started in Material Physics, later transferred into Physics.从材料物理入学，中途转入物理学。
- Undergraduate research on general relativity and orbital mechanics.本科研究方向是广义相对论与天体轨道力学。
Awards荣誉
-
2021 Top Ten Students / Star of Science and Technology十佳大学生 / 科技之星
Inner Mongolia Normal University内蒙古师范大学
-
2019 "Challenge Cup" — Inner Mongolia Autonomous Region Special Award「挑战杯」内蒙古自治区 特等奖
Inner Mongolia Autonomous Region内蒙古自治区
Projects项目
-
Orbital dynamics in strongly curved spacetime强引力时空下的轨道动力学
2020 – 2022 · funded student research project (College Students’ Innovative Training Program of the Inner Mongolia Education Department). Produced the first-author Chinese Physics B paper on Mercury’s gravitation-induced orbital shrinkage.2020 – 2022 · 内蒙古自治区教育厅资助的大学生创新训练计划。这一项目最终产出了发表于 Chinese Physics B 的水星轨道引力收缩论文（第一作者）。
-
Preparation of Cu-Zn-Sn-S (CZTS) thin films铜锌锡硫（CZTS）薄膜的制备
2018 · National College Students’ Innovation and Entrepreneurship Training Program (funded by the Ministry of Education of China). Advisor: Prof. Yang Yanchun. Materials-physics-era project from the first year of undergrad, before switching into physics.2018 · 国家级大学生创新创业训练计划（教育部资助）。指导老师：杨艳春 老师。本科入学第一年、转入物理学之前在材料物理方向做的项目。
Publications论文
-
2026 Neural Non-Equilibrium Hamiltonian Monte Carlo for Corrected Boltzmann Sampling
arXiv
Learned stochastic Hamiltonian paths with recorded-work correction: path-SNIS/path-IMH plus a proved shared-bridge round-trip Metropolis kernel. Tested on analytic many-well and finite-volume 2D phi^4 targets; molecular and compact U(1) results retain explicit feasibility boundaries.用记录功修正可学习的随机 Hamiltonian 路径：包括 path-SNIS / path-IMH，以及经过证明的共享桥接点 round-trip Metropolis 核。已在解析多阱与有限体积二维 phi^4 上测试；分子和紧致 U(1) 结果保留明确的可行性边界。
-
2026 Stochastic Path Sampler For Lattice Field Theory
arXiv
Data-free neural sampler based on nonequilibrium thermodynamics and stochastic quantization; matches HMC quality with much shorter autocorrelation in the pseudocritical region of 2D phi^4. Co-first author with S. Chen.基于非平衡热力学与随机量子化的数据无关神经采样器；在 2D φ⁴ 伪临界区，质量与 HMC 相当，自相关时间显著更短。与 S. Chen 共同一作。
-
2026 Operator Spectroscopy of Trained Lattice Samplers
arXiv
Generative Boltzmann samplers still hit scaling, unbalanced mode weights, and mode collapse. Gauge fields make this worse: an equivariant architecture is almost required, so training becomes expensive. This work therefore reads off the analytic operator bases a trained network actually uses, and asks whether those bases can be built into a smaller architecture that trains more cheaply.生成式玻尔兹曼采样器仍然会碰到标度问题、模态权重失衡和模态崩溃。规范场让这件事更糟：几乎必须用等变架构，训练因此非常贵。本文由此读出训练后网络实际用到的解析算符基，并问这些基能不能做进一个更小的架构，从而更便宜地训练。
-
2025 Variational Autoregressive Networks Applied to $\phi^4$ Field Theory Systems
arXiv
RL-style data-free training of variational autoregressive networks for the Ising model and 2D phi^4; single-site and block Metropolis-Hastings corrections reduce residual bias of pure VAN sampling.用强化学习方式数据无关地训练变分自回归网络，应用于 Ising 模型与 2D φ⁴；单点 / 块 Metropolis-Hastings 修正显著降低纯 VAN 采样的剩余偏差。
-
2020 Gravitation induced shrinkage of Mercury's orbit
Chinese Physics B 29 (2020) 109501
General-relativistic gravitational contribution to the secular shrinkage of Mercury’s orbit.水星轨道长期收缩的广义相对论引力修正贡献。