General Lab Information

Hsin Yi Lin

Computational Science

Hsin Yi Lin

Brookhaven National Laboratory

Computational Science
P.O. Box 5000
Upton, NY 11973-5000

Hsin-Yi (Lucy) Lin is a Research Scientist in Computing and Data Sciences at Brookhaven National Laboratory, working on quantum computing and quantum machine learning (QML). Her research develops variational quantum circuits and quantum–classical hybrid models, with a focus on learnable measurement design (adaptive non-local observables), interpretability, and practical quantum advantage. Her broader work spans large-scale deep learning—Transformers/LLMs, diffusion models, and optimal transport—applied to speech and language, climate and geophysical dynamics, and ecology.

Expertise | Research | Education | Appointments | Publications | Highlights

Expertise

  • Quantum machine learning and variational quantum circuits
  • Hybrid quantum–classical algorithms
  • Large language models and foundation models
  • Trustworthy and interpretable AI
  • Scientific machine learning and dynamical systems
  • Applied analysis and partial differential equations

Research Activities

  • Quantum machine learning: variational quantum circuits, hybrid models, and learnable (adaptive non-local) observables for expressive, efficient quantum neural networks.
  • Quantum learning applications: quantum reinforcement learning and quantum super-resolution.
  • Trustworthy and interpretable AI: model interpretability for classical and quantum networks (e.g., Quantum Gradient Class Activation Map).
  • Foundation models: LLMs, retrieval-augmented generation, diffusion and contrastive learning.
  • Scientific machine learning: forecasting chaotic geophysical systems, data assimilation, and transport phenomena.

Education

  • Ph.D. in Mathematics, University of Maryland, College Park, 2020
  • M.S. in Mathematics, National Tsing Hua University, Taiwan
  • B.S. in Physics and Mathematics (dual major), National Tsing Hua University, Taiwan

Professional Appointments

  • 2026–present: Research Scientist, Brookhaven National Laboratory
  • 2023–2025: Assistant Professor, Department of Mathematics & Computer Science, Seton Hall University

Selected Publications

  • Lin H-Y, Chang Y-C, Chien J-T (2026) Variational Dehallucination via Retrieval-Augmented Prompt Learning. IEEE Transactions on Audio, Speech and Language Processing 34:259–271. https://doi.org/10.1109/taslpro.2025.3646477
  • Lin H-Y, Tseng H-H, Chen SY-C, Yoo S (2025) Adaptive Non-Local Observable on Quantum Neural Networks. 2025 IEEE International Conference on Quantum Computing and Engineering (QCE), 1884–1893. https://doi.org/10.1109/QCE65121.2025.00206
  • Lin H-Y, Chen SY-C, Tseng H-H, Yoo S (2025) Quantum Reinforcement Learning by Adaptive Non-Local Observables. 2025 IEEE International Conference on Quantum Computing and Engineering (QCE) 241–246. https://doi.org/10.1109/qce65121.2025.10326
  • Lin H-Y, Tseng H-H, Chen SY-C, Yoo S (2024) Quantum Gradient Class Activation Map for Model Interpretability. 2024 IEEE Workshop on Signal Processing Systems (SiPS) 165–170. https://doi.org/10.1109/sips62058.2024.00037
  • Lin H-Y, Tseng H-H, Chien J-T (2024) On the Attractive and Repulsive Forces of Generalized Stochastic Neighbor Embedding With Alpha-Divergence. IEEE Access 12:90380–90394. https://doi.org/10.1109/access.2024.3420425
  • Tseng H-H, Lin H-Y, Hung K-H, Tsao Y (2023) Interpretations of Domain Adaptations via Layer Variational Analysis. International Conference on Learning Representations (ICLR). https://arxiv.org/abs/2302.01798
  • Jabin P-E, Lin H-Y, Tadmor E (2022) Commutator method for averaging lemmas. Analysis & PDE 15:1561–1584. https://doi.org/10.2140/apde.2022.15.1561
  • Penny SG, Smith TA, Chen T-C, et al (2022) Integrating Recurrent Neural Networks With Data Assimilation for Scalable Data-Driven State Estimation. Journal of Advances in Modeling Earth Systems 14:e2021MS002843. https://doi.org/10.1029/2021ms002843
  • Lin H-Y, Tseng H-H, Lu X, Tsao Y (2021) Unsupervised Noise Adaptive Speech Enhancement by Discriminator-Constrained Optimal Transport. Advances in Neural Information Processing Systems (NeurIPS) 34. https://proceedings.neurips.cc/paper_files/paper/2021/file/a5c7b30fb632c92feb59154517223dc9-Paper.pdf
  • Lin H-Y, Fagan WF, Jabin P-E (2021) Memory-driven movement model for periodic migrations. Journal of Theoretical Biology 508:110486. https://doi.org/10.1016/j.jtbi.2020.110486

Research Highlights

My research builds on mathematical analysis and extends it to modeling, machine learning, and quantum algorithms.

  • Partial differential equations. With P.-E. Jabin and E. Tadmor, I developed a commutator method for velocity averaging lemmas in kinetic transport equations, linking them to dispersive smoothing (Analysis & PDE, 2022).
  • Mathematical modeling. build differential-equation and stochastic models of real systems, such as memory-driven animal migration (J. Theoretical Biology, 2021) and disease spread.
  • Machine learning. apply analysis and optimization tools to deep learning, including optimal transport for speech enhancement (NeurIPS 2021), variational analysis of domain adaptation (ICLR 2023), and neural networks with data assimilation for chaotic systems (JAMES, 2022).
  • Quantum algorithms. Inspired by the Heisenberg picture, I introduced adaptive non-local observables: trainable measurements that make variational quantum circuits more expressive without adding depth (IEEE QCE 2025). My collaborators and I have applied them to quantum reinforcement learning and super-resolution, and developed tools for interpreting hybrid quantum models (IEEE SiPS 2024).
Hsin Yi Lin

Brookhaven National Laboratory

Computational Science
P.O. Box 5000
Upton, NY 11973-5000

Hsin Yi's Links