(Louie) Hong Yao

profilejpg.jpg

I am a statistical physicist, and my work is guided by Philip W. Anderson’s idea that “more is different.” When many simple components interact, new laws and structures can emerge that are not obvious from the parts alone. In statistical mechanics, I focus on phase transitions, universality, and symmetries, using them to understand how macroscopic behavior becomes insensitive to microscopic details. In quantum computing and information, I use statistical-physics ideas to study how entanglement, correlations, and collective dynamics emerge in many-body quantum systems and how these structures can be used for computation and information processing. In machine learning, I aim to understand representations and learning dynamics in large models, treating them as evolving, high-dimensional systems shaped by optimization, data, and scale.

I earned my Ph.D. in theoretical physics at Virginia Tech (2018–2023) under Prof. Uwe C. Täuber, and my B.Sc. in Physics from Nankai University (2014–2018). I currently work as a predictive modeler at The Cincinnati Insurance Company.

News

Sep 27, 2026 New preprint on arXiv: MinkowskiPE: Minkowski Positional Encoding for Spatiotemporal Perception.
Sep 22, 2026 New preprint on arXiv: The Linear Representation Hypothesis Needs a Group Action.
Jun 08, 2026 Paper accepted at TMLR: JE-IRT: A Geometric Lens on LLM Abilities through Joint Embedding Item Response Theory.
May 28, 2026 New preprint on arXiv: Cross-Lingual Steering for Figurative Language Generation.
Apr 11, 2026 New preprint on arXiv: A Minimal Model of Representation Collapse: Frustration, Stop-Gradient, and Dynamics.

Selected Publications

  1. arXiv
    MinkowskiPE: Minkowski Positional Encoding for Spatiotemporal Perception
    Y. Li, L. H. Yao, T. Shi, H. Cao, H. Hao, Z. Zhao, and S. Liu
    arXiv preprint arXiv:2609.33804, 2026
    arXiv
  2. arXiv
    The Linear Representation Hypothesis Needs a Group Action
    L. H. Yao, Y. Li, and S. Liu
    arXiv preprint arXiv:2609.27158, 2026
    arXiv
  3. TMLR
    JE-IRT: A Geometric Lens on LLM Abilities through Joint Embedding Item Response Theory
    L. H. Yao, N. Jarvis, T. Zhan, S. Ghosh, L. Liu, and T. Jiang
    Transactions on Machine Learning Research, 2026
    arXiv HTML
  4. arXiv
    Cross-Lingual Steering for Figurative Language Generation
    L. Liu, T. Zhan, L. H. Yao, S. Ghosh, and T. Jiang
    arXiv preprint arXiv:2605.30443, 2026
    arXiv
  5. arXiv
    A Minimal Model of Representation Collapse: Frustration, Stop-Gradient, and Dynamics
    L. H. Yao, Y. Li, and S. Liu
    arXiv preprint arXiv:2604.09979, 2026
    arXiv
  6. ACL
    Rhetorical Questions in LLM Representations: A Linear Probing Study
    L. H. Yao, V. Anand, Y. Zhuang, and T. Jiang
    In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics, 2026
    arXiv HTML
  7. EACL
    Towards Robust Evaluation of Visual Activity Recognition: Resolving Verb Ambiguity with Sense Clustering
    L. H. Yao, N. Jarvis, and T. Jiang
    In Findings of the Association for Computational Linguistics: EACL, 2026
    arXiv HTML
  8. Quantum
    Stochastic Resetting in Discrete-time Quantum Dynamics: Steady States and Correlations in Few-qubit Systems
    S. Wald, L. H. Yao, T. Platini, C. Hooley, and F. Carollo
    Quantum, 2025
    arXiv HTML
  9. Phys. Rev. E
    Coined Quantum Walks on the Line: Disorder, Entanglement and Localization
    L. H. Yao and S. Wald
    Physical Review E, 2023
    arXiv HTML
  10. Phys. Rev. E
    Critical Dynamics of the Antiferromagnetic O(3) Nonlinear Sigma Model with Conserved Magnetization
    L. H. Yao and U. C. Täuber
    Physical Review E, 2022
    arXiv HTML