2026
About
I am a Ph.D. student in the Artificial Intelligence Thrust, Information Hub, at The Hong Kong University of Science and Technology (Guangzhou), advised by Prof. Yao Shu. Previously, I received my master’s and bachelor’s degrees in Physics from Sun Yat-sen University, where I was advised by Prof. Haiping Huang.
My research focuses on developing theoretically grounded and practically efficient algorithms for modern AI systems. I study variance-reduced zeroth-order optimization for black-box objectives and on-device LLM fine-tuning; efficient LLM post-training in online and offline settings; and post-training quantization and edge deployment, balancing performance, efficiency, and privacy.
Selected Publications
Revisiting Zeroth-Order Hessian Approximation: A Single-Step Policy Optimization Lens
A policy-optimization view of zeroth-order curvature estimation, with variance reduction through baseline selection and query reuse.
BibTeX
@inproceedings{qiu2026zovh,
title = {Revisiting Zeroth-Order Hessian Approximation: A Single-Step Policy Optimization Lens},
author = {Junbin Qiu and Zhaowei Hong and Renzhe Xu and Yao Shu},
booktitle = {International Conference on Machine Learning (ICML)},
month = {jul},
volume = {306},
pages = {102036--102072},
publisher = {PMLR},
year = {2026},
url = {https://proceedings.mlr.press/v306/qiu26h.html}
}2025
Zeroth-Order Optimization is Secretly Single-Step Policy Optimization
Connects finite-difference zeroth-order optimization with single-step policy optimization, motivating an averaged baseline and query reuse for variance reduction.
BibTeX
@inproceedings{qiu2025zoar,
title = {Zeroth-Order Optimization is Secretly Single-Step Policy Optimization},
author = {Junbin Qiu and Zhengpeng Xie and Xiangda Yan and Yongjie Yang and Yao Shu},
booktitle = {ICML 2025 Workshop},
month = {jun},
year = {2025},
url = {https://arxiv.org/abs/2506.14460}
}Education
The Hong Kong University of Science and Technology (Guangzhou)
2025 – presentPh.D. student in Artificial Intelligence
Artificial Intelligence Thrust, Information Hub
Sun Yat-sen University
2021 – 2024Master’s degree in Statistical Physics
Sun Yat-sen University
2017 – 2021Bachelor’s degree in Physics
Publications
8 matching publications
2026
APEX: Accuracy Projection from Verifier-Labeled Experience for Offline RLVR
Revisiting Zeroth-Order Hessian Approximation: A Single-Step Policy Optimization Lens
BibTeX
@inproceedings{qiu2026zovh,
title = {Revisiting Zeroth-Order Hessian Approximation: A Single-Step Policy Optimization Lens},
author = {Junbin Qiu and Zhaowei Hong and Renzhe Xu and Yao Shu},
booktitle = {International Conference on Machine Learning (ICML)},
month = {jul},
volume = {306},
pages = {102036--102072},
publisher = {PMLR},
year = {2026},
url = {https://proceedings.mlr.press/v306/qiu26h.html}
}2025
Zeroth-Order Optimization is Secretly Single-Step Policy Optimization
BibTeX
@inproceedings{qiu2025zoar,
title = {Zeroth-Order Optimization is Secretly Single-Step Policy Optimization},
author = {Junbin Qiu and Zhengpeng Xie and Xiangda Yan and Yongjie Yang and Yao Shu},
booktitle = {ICML 2025 Workshop},
month = {jun},
year = {2025},
url = {https://arxiv.org/abs/2506.14460}
}An optimization-based equilibrium measure describing fixed points of non-equilibrium dynamics: application to the edge of chaos
BibTeX
@article{qiu2025equilibrium,
title = {An optimization-based equilibrium measure describing fixed points of non-equilibrium dynamics: application to the edge of chaos},
author = {Junbin Qiu and Haiping Huang},
journal = {Communications in Theoretical Physics},
month = {mar},
volume = {77},
number = {3},
pages = {035601},
doi = {10.1088/1572-9494/ad8126},
year = {2025},
url = {https://ctp.itp.ac.cn/EN/10.1088/1572-9494/ad8126}
}2024
Meta predictive learning model of languages in neural circuits
BibTeX
@article{li2024metapredictive,
title = {Meta predictive learning model of languages in neural circuits},
author = {Chan Li and Junbin Qiu and Haiping Huang},
journal = {Physical Review E},
volume = {109},
pages = {044309},
doi = {10.1103/PhysRevE.109.044309},
year = {2024},
url = {https://journals.aps.org/pre/abstract/10.1103/PhysRevE.109.044309}
}2022
Equivalence between belief propagation instability and transition to replica symmetry breaking in perceptron learning systems
BibTeX
@article{zhao2022perceptron,
title = {Equivalence between belief propagation instability and transition to replica symmetry breaking in perceptron learning systems},
author = {Yang Zhao and Junbin Qiu and Mingshan Xie and Haiping Huang},
journal = {Physical Review Research},
volume = {4},
pages = {023023},
doi = {10.1103/PhysRevResearch.4.023023},
year = {2022},
url = {https://journals.aps.org/prresearch/abstract/10.1103/PhysRevResearch.4.023023}
}Activities
Talk
A language model based on free energy minimization
Contact
I welcome research discussions and collaborations around theoretical machine learning, zeroth-order optimization, efficient LLM post-training, and model deployment.