Prior-Guided Symbolic Regression: Towards Scientific Consistency in Equation Discovery
arXiv, 2026
About Me
你好，我是肖劲，目前是国防科技大学计算机科学与技术专业的四年级直博生，于 2023 年 9 月入学，预计于2028 年 6 月毕业。本科期间，我就读于中山大学信息与计算科学专业（2019–2023）。
我的研究兴趣包括科学计算、神经算子、物理信息学习，以及面向复杂科学与工程问题的机器学习方法。我关注如何利用学习型方法求解偏微分方程、生成结构化网格，并为科学计算任务构建高效的代理模型。
我目前正在寻找研究实习机会，方向包括科学计算、神经算子、物理信息学习及相关人工智能领域。也欢迎开展学术交流与合作，请通过 xiaoj10@nudt.edu.cn 与我联系。
I am Jing Xiao, a fourth-year direct-entry Ph.D. student in Computer Science and Technology at the National University of Defense Technology. I began my Ph.D. studies in September 2023 and expect to graduate in June 2028. Previously, I obtained my bachelor's degree in Information and Computing Science from Sun Yat-sen University in 2023.
My research interests include scientific computing, neural operators, physics-informed learning, and machine-learning methods for complex scientific and engineering problems. I am particularly interested in learning-based approaches for solving partial differential equations, generating structured meshes, and building efficient surrogate models for scientific computing.
I am currently seeking research internship opportunities in scientific computing, neural operators, physics-informed learning, and related areas of AI. I also welcome academic discussions and research collaborations. Feel free to reach out at xiaoj10@nudt.edu.cn.
News
- 2026.09首次向 arXiv 提交预印本《Gradient-Update Mismatch: Rethinking Conflict-Free Training of Physics-Informed Neural Networks》。
- 2026.08论文《LDNO: A low-power dynamic neural operator inspired by liquid state machines for solving partial differential equations》发表于 Neurocomputing（CCF-C，JCR Q1）。
- 2026.06论文《Physics-informed residual learning with low-rank adaptation for unsupervised mesh generation》发表于 Computer-Aided Geometric Design（CCF-B，JCR Q2）。
- 2026.03论文《Learning to Generate Structured Meshes with In-Context: Toward Generalization in Mesh Generation》在线发表于 AAAI 2026（CCF-A） 会议论文集。
- 2026.02论文《MeshONet: A Generalizable and Efficient Operator Learning Method for Structured Mesh Generation》在线发表于 Neural Networks（CCF-B，JCR Q1）。
- 2026.02首次向 arXiv 提交预印本《Prior-Guided Symbolic Regression: Towards Scientific Consistency in Equation Discovery》。
- 2025.12论文《PI-MeshONet: A generalizable and self-supervised method for structured mesh generation》在线发表于 Knowledge-Based Systems（CCF-C，JCR Q1）。
- 2025.06论文《Dual-Spectral Neural Operator for Solving Partial Differential Equations》发表于 IJCNN 2025（CCF-C） 会议论文集。
- 2025.03论文《A Physics-Informed Generative Adversarial Network for Advancing Solutions in Ocean Acoustics》在线发表于 Physics of Fluids（JCR Q1）。
- 2023.09开始在 国防科技大学 攻读计算机科学与技术专业博士学位。
- Sep 2026First submitted “Gradient-Update Mismatch: Rethinking Conflict-Free Training of Physics-Informed Neural Networks” to arXiv.
- Aug 2026Our paper “LDNO: A low-power dynamic neural operator inspired by liquid state machines for solving partial differential equations” was published in Neurocomputing (CCF-C, JCR Q1).
- Jun 2026Our paper “Physics-informed residual learning with low-rank adaptation for unsupervised mesh generation” was published in Computer-Aided Geometric Design (CCF-B, JCR Q2).
- Mar 2026Our paper “Learning to Generate Structured Meshes with In-Context: Toward Generalization in Mesh Generation” was published online in the proceedings of AAAI 2026 (CCF-A).
- Feb 2026Our paper “MeshONet: A Generalizable and Efficient Operator Learning Method for Structured Mesh Generation” was published online in Neural Networks (CCF-B, JCR Q1).
- Feb 2026First submitted “Prior-Guided Symbolic Regression: Towards Scientific Consistency in Equation Discovery” to arXiv.
- Dec 2025Our paper “PI-MeshONet: A generalizable and self-supervised method for structured mesh generation” was published online in Knowledge-Based Systems (CCF-C, JCR Q1).
- Jun 2025Our paper “Dual-Spectral Neural Operator for Solving Partial Differential Equations” was published in the proceedings of IJCNN 2025 (CCF-C).
- Mar 2025Our paper “A Physics-Informed Generative Adversarial Network for Advancing Solutions in Ocean Acoustics” was published online in Physics of Fluids (JCR Q1).
- Sep 2023Started my Ph.D. in Computer Science and Technology at the National University of Defense Technology.
My Research
Learning to Generate Structured Meshes with In-Context: Toward Generalization in Mesh Generation
AAAI Conference on Artificial Intelligence (CCF-A), 2026
PI-MeshONet: A generalizable and self-supervised method for structured mesh generation
Knowledge-Based Systems (CCF-C, JCR Q1), 2026
MeshONet: A Generalizable and Efficient Operator Learning Method for Structured Mesh Generation
Neural Networks (CCF-B, JCR Q1), 2026
Dual-Spectral Neural Operator for Solving Partial Differential Equations
International Joint Conference on Neural Networks (CCF-C), 2025
LDNO: A low-power dynamic neural operator inspired by liquid state machines for solving partial differential equations
Neurocomputing (CCF-C, JCR Q1), 2026
