Skip to content
View luckyjackluo's full-sized avatar
  • UC San Diego
  • San Diego

Block or report luckyjackluo

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
luckyjackluo/README.md

Zhishang (Jack) Luo — Research Portfolio

PhD Student at Halıcıoğlu Data Science Institute, UC San Diego


About Me

I am a graph learning and combinatorial optimization researcher pursuing a PhD under Prof. Yusu Wang, University of California San Diego.


Research Interests

  • Graph & Hypergraph Neural Networks (DE-HNN for circuit netlist representation)
  • Differentiable Optimization over Permutations (with rounding guarantees)
  • ML Infrastructure & Many-body Potentials (MB-Fit framework)
  • Neuroscience & Memory Consolidation (verbal long-term memory via fMRI)

Selected Publications

  1. DE-HNN: An Effective Neural Model for Circuit Netlist Representation

    • AISTATS 2024, CoRR 2024 :contentReference[oaicite:2]{index=2}
    • Introduces a Directional Equivariant Hypergraph Neural Network, modeling netlists as directed hypergraphs. Achieves universal approximation under permutation equivariance/invariance and outperforms SOTA models in place-and-route outcome prediction :contentReference[oaicite:3]{index=3}.
  2. Differentiable Extensions with Rounding Guarantees for Combinatorial Optimization over Permutations

    • CoRR 2024 :contentReference[oaicite:4]{index=4}
    • Presents the Birkhoff Extension (BE): a differentiable, continuous, polynomial-time computable extension from permutations to doubly stochastic matrices. Guarantees efficient rounding without loss of objective value and enables gradient-based optimization for permutation objectives :contentReference[oaicite:5]{index=5}.
  3. Human Brain Activity & Functional Connectivity for Verbal Long-Term Memory Consolidation (1-Month fMRI Study)

    • Frontiers in Human Neuroscience, February 2024 :contentReference[oaicite:6]{index=6}
    • Identifies brain activity and connectivity shifts (e.g., hippocampus, vmPFC) over memory aging from 1 hour to 1 month, supporting standard systems consolidation theories :contentReference[oaicite:7]{index=7}.
  4. “Short Blanket” Dilemma for Neural Network Potentials of Water

    • J. Chemical Physics, February 2023 :contentReference[oaicite:8]{index=8}
    • Highlights the trade-off: DNN potentials can replicate MB-pol liquid water accuracy but fail on vapor-liquid equilibrium due to poor many-body interaction learning.

Fellowship & Awards

  • 2024 Qualcomm Innovation Fellowship for the project "Explainable Graph Learning for Property Prediction on Netlist Representations" :contentReference[oaicite:9]{index=9}.

Repository Structure & Usage

Popular repositories Loading

  1. TopoGeoNet TopoGeoNet Public

    This is the repo for TopoGeoNet: A scalable learning framework for large-scale spatial graph.

    Python 1

  2. luckyjackluo luckyjackluo Public

    Config files for my GitHub profile.

  3. Gopher Gopher Public

    Interpretable Data-Based Explanations for Fairness Debugging

    Jupyter Notebook

  4. MB-Fit MB-Fit Public

    Optimization and further development of MB_Fit software. This project is a lab project of The Paesani Research Group.

    C++

  5. AGOP-GNN AGOP-GNN Public

    The repository for AGOP on GNN.

    Jupyter Notebook

  6. BE-for-Combinatorial-Optimization BE-for-Combinatorial-Optimization Public

    This is the repo for using Extension to solve combinatorial optimization problems.

    Jupyter Notebook